-
+
diff --git a/backend/alembic/versions/a3bfd0d64902_add_chosen_assistants_to_user_table.py b/backend/alembic/versions/a3bfd0d64902_add_chosen_assistants_to_user_table.py
new file mode 100644
index 00000000000..89439adb680
--- /dev/null
+++ b/backend/alembic/versions/a3bfd0d64902_add_chosen_assistants_to_user_table.py
@@ -0,0 +1,27 @@
+"""Add chosen_assistants to User table
+
+Revision ID: a3bfd0d64902
+Revises: ec85f2b3c544
+Create Date: 2024-05-26 17:22:24.834741
+
+"""
+from alembic import op
+import sqlalchemy as sa
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision = "a3bfd0d64902"
+down_revision = "ec85f2b3c544"
+branch_labels: None = None
+depends_on: None = None
+
+
+def upgrade() -> None:
+ op.add_column(
+ "user",
+ sa.Column("chosen_assistants", postgresql.ARRAY(sa.Integer()), nullable=True),
+ )
+
+
+def downgrade() -> None:
+ op.drop_column("user", "chosen_assistants")
diff --git a/backend/alembic/versions/b85f02ec1308_fix_file_type_migration.py b/backend/alembic/versions/b85f02ec1308_fix_file_type_migration.py
new file mode 100644
index 00000000000..ac17670b703
--- /dev/null
+++ b/backend/alembic/versions/b85f02ec1308_fix_file_type_migration.py
@@ -0,0 +1,28 @@
+"""fix-file-type-migration
+
+Revision ID: b85f02ec1308
+Revises: a3bfd0d64902
+Create Date: 2024-05-31 18:09:26.658164
+
+"""
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision = "b85f02ec1308"
+down_revision = "a3bfd0d64902"
+branch_labels: None = None
+depends_on: None = None
+
+
+def upgrade() -> None:
+ op.execute(
+ """
+ UPDATE file_store
+ SET file_origin = UPPER(file_origin)
+ """
+ )
+
+
+def downgrade() -> None:
+ # Let's not break anything on purpose :)
+ pass
diff --git a/backend/danswer/auth/noauth_user.py b/backend/danswer/auth/noauth_user.py
new file mode 100644
index 00000000000..4744c4a6488
--- /dev/null
+++ b/backend/danswer/auth/noauth_user.py
@@ -0,0 +1,40 @@
+from collections.abc import Mapping
+from typing import Any
+from typing import cast
+
+from danswer.auth.schemas import UserRole
+from danswer.dynamic_configs.store import ConfigNotFoundError
+from danswer.dynamic_configs.store import DynamicConfigStore
+from danswer.server.manage.models import UserInfo
+from danswer.server.manage.models import UserPreferences
+
+
+NO_AUTH_USER_PREFERENCES_KEY = "no_auth_user_preferences"
+
+
+def set_no_auth_user_preferences(
+ store: DynamicConfigStore, preferences: UserPreferences
+) -> None:
+ store.store(NO_AUTH_USER_PREFERENCES_KEY, preferences.dict())
+
+
+def load_no_auth_user_preferences(store: DynamicConfigStore) -> UserPreferences:
+ try:
+ preferences_data = cast(
+ Mapping[str, Any], store.load(NO_AUTH_USER_PREFERENCES_KEY)
+ )
+ return UserPreferences(**preferences_data)
+ except ConfigNotFoundError:
+ return UserPreferences(chosen_assistants=None)
+
+
+def fetch_no_auth_user(store: DynamicConfigStore) -> UserInfo:
+ return UserInfo(
+ id="__no_auth_user__",
+ email="anonymous@danswer.ai",
+ is_active=True,
+ is_superuser=False,
+ is_verified=True,
+ role=UserRole.ADMIN,
+ preferences=load_no_auth_user_preferences(store),
+ )
diff --git a/backend/danswer/chat/load_yamls.py b/backend/danswer/chat/load_yamls.py
index abb10461a9f..4036730a9ab 100644
--- a/backend/danswer/chat/load_yamls.py
+++ b/backend/danswer/chat/load_yamls.py
@@ -6,13 +6,13 @@
from danswer.configs.chat_configs import MAX_CHUNKS_FED_TO_CHAT
from danswer.configs.chat_configs import PERSONAS_YAML
from danswer.configs.chat_configs import PROMPTS_YAML
-from danswer.db.chat import get_prompt_by_name
-from danswer.db.chat import upsert_persona
-from danswer.db.chat import upsert_prompt
from danswer.db.document_set import get_or_create_document_set_by_name
from danswer.db.engine import get_sqlalchemy_engine
from danswer.db.models import DocumentSet as DocumentSetDBModel
from danswer.db.models import Prompt as PromptDBModel
+from danswer.db.persona import get_prompt_by_name
+from danswer.db.persona import upsert_persona
+from danswer.db.persona import upsert_prompt
from danswer.search.enums import RecencyBiasSetting
diff --git a/backend/danswer/chat/process_message.py b/backend/danswer/chat/process_message.py
index 06ac84fa061..7ba7d991aa2 100644
--- a/backend/danswer/chat/process_message.py
+++ b/backend/danswer/chat/process_message.py
@@ -14,6 +14,7 @@
from danswer.chat.models import QADocsResponse
from danswer.chat.models import StreamingError
from danswer.configs.chat_configs import CHAT_TARGET_CHUNK_PERCENTAGE
+from danswer.configs.chat_configs import DISABLE_LLM_CHOOSE_SEARCH
from danswer.configs.chat_configs import MAX_CHUNKS_FED_TO_CHAT
from danswer.configs.constants import MessageType
from danswer.db.chat import attach_files_to_chat_message
@@ -48,6 +49,8 @@
from danswer.search.enums import OptionalSearchSetting
from danswer.search.retrieval.search_runner import inference_documents_from_ids
from danswer.search.utils import chunks_or_sections_to_search_docs
+from danswer.search.utils import dedupe_documents
+from danswer.search.utils import drop_llm_indices
from danswer.server.query_and_chat.models import ChatMessageDetail
from danswer.server.query_and_chat.models import CreateChatMessageRequest
from danswer.server.utils import get_json_line
@@ -94,14 +97,21 @@ def _handle_search_tool_response_summary(
packet: ToolResponse,
db_session: Session,
selected_search_docs: list[DbSearchDoc] | None,
-) -> tuple[QADocsResponse, list[DbSearchDoc]]:
+ dedupe_docs: bool = False,
+) -> tuple[QADocsResponse, list[DbSearchDoc], list[int] | None]:
response_sumary = cast(SearchResponseSummary, packet.response)
+ dropped_inds = None
if not selected_search_docs:
top_docs = chunks_or_sections_to_search_docs(response_sumary.top_sections)
+
+ deduped_docs = top_docs
+ if dedupe_docs:
+ deduped_docs, dropped_inds = dedupe_documents(top_docs)
+
reference_db_search_docs = [
- create_db_search_doc(server_search_doc=top_doc, db_session=db_session)
- for top_doc in top_docs
+ create_db_search_doc(server_search_doc=doc, db_session=db_session)
+ for doc in deduped_docs
]
else:
reference_db_search_docs = selected_search_docs
@@ -121,12 +131,17 @@ def _handle_search_tool_response_summary(
recency_bias_multiplier=response_sumary.recency_bias_multiplier,
),
reference_db_search_docs,
+ dropped_inds,
)
def _check_should_force_search(
new_msg_req: CreateChatMessageRequest,
) -> ForceUseTool | None:
+ # If files are already provided, don't run the search tool
+ if new_msg_req.file_descriptors:
+ return None
+
if (
new_msg_req.query_override
or (
@@ -134,6 +149,7 @@ def _check_should_force_search(
and new_msg_req.retrieval_options.run_search == OptionalSearchSetting.ALWAYS
)
or new_msg_req.search_doc_ids
+ or DISABLE_LLM_CHOOSE_SEARCH
):
args = (
{"query": new_msg_req.query_override}
@@ -267,8 +283,8 @@ def stream_chat_message_objects(
"Be sure to update the chat pointers before calling this."
)
- # Save now to save the latest chat message
- db_session.commit()
+ # NOTE: do not commit user message - it will be committed when the
+ # assistant message is successfully generated
else:
# re-create linear history of messages
final_msg, history_msgs = create_chat_chain(
@@ -298,6 +314,7 @@ def stream_chat_message_objects(
new_file.to_file_descriptor() for new_file in latest_query_files
],
db_session=db_session,
+ commit=False,
)
selected_db_search_docs = None
@@ -356,7 +373,7 @@ def stream_chat_message_objects(
# error=,
# reference_docs=,
db_session=db_session,
- commit=True,
+ commit=False,
)
if not final_msg.prompt:
@@ -386,7 +403,7 @@ def stream_chat_message_objects(
search_tool: SearchTool | None = None
tools: list[Tool] = []
for tool_cls in persona_tool_classes:
- if tool_cls.__name__ == SearchTool.__name__:
+ if tool_cls.__name__ == SearchTool.__name__ and not latest_query_files:
search_tool = SearchTool(
db_session=db_session,
user=user,
@@ -445,25 +462,44 @@ def stream_chat_message_objects(
PreviousMessage.from_chat_message(msg, files) for msg in history_msgs
],
tools=tools,
- force_use_tool=_check_should_force_search(new_msg_req),
+ force_use_tool=(
+ _check_should_force_search(new_msg_req) if search_tool else None
+ ),
)
reference_db_search_docs = None
qa_docs_response = None
ai_message_files = None # any files to associate with the AI message e.g. dall-e generated images
+ dropped_indices = None
for packet in answer.processed_streamed_output:
if isinstance(packet, ToolResponse):
if packet.id == SEARCH_RESPONSE_SUMMARY_ID:
(
qa_docs_response,
reference_db_search_docs,
+ dropped_indices,
) = _handle_search_tool_response_summary(
- packet, db_session, selected_db_search_docs
+ packet=packet,
+ db_session=db_session,
+ selected_search_docs=selected_db_search_docs,
+ # Deduping happens at the last step to avoid harming quality by dropping content early on
+ dedupe_docs=retrieval_options.dedupe_docs
+ if retrieval_options
+ else False,
)
yield qa_docs_response
elif packet.id == SECTION_RELEVANCE_LIST_ID:
+ chunk_indices = packet.response
+
+ if reference_db_search_docs is not None and dropped_indices:
+ chunk_indices = drop_llm_indices(
+ llm_indices=chunk_indices,
+ search_docs=reference_db_search_docs,
+ dropped_indices=dropped_indices,
+ )
+
yield LLMRelevanceFilterResponse(
- relevant_chunk_indices=packet.response
+ relevant_chunk_indices=chunk_indices
)
elif packet.id == IMAGE_GENERATION_RESPONSE_ID:
img_generation_response = cast(
@@ -485,11 +521,19 @@ def stream_chat_message_objects(
yield cast(ChatPacket, packet)
except Exception as e:
- logger.exception(e)
+ logger.exception("Failed to process chat message")
+
+ # Don't leak the API key
+ error_msg = str(e)
+ if llm.config.api_key and llm.config.api_key.lower() in error_msg.lower():
+ error_msg = (
+ f"LLM failed to respond. Invalid API "
+ f"key error from '{llm.config.model_provider}'."
+ )
- # Frontend will erase whatever answer and show this instead
- # This will be the issue 99% of the time
- yield StreamingError(error="LLM failed to respond, have you set your API key?")
+ yield StreamingError(error=error_msg)
+ # Cancel the transaction so that no messages are saved
+ db_session.rollback()
return
# Post-LLM answer processing
@@ -513,6 +557,7 @@ def stream_chat_message_objects(
citations=db_citations,
error=None,
)
+ db_session.commit() # actually save user / assistant message
msg_detail_response = translate_db_message_to_chat_message_detail(
gen_ai_response_message
diff --git a/backend/danswer/chat/prompts.yaml b/backend/danswer/chat/prompts.yaml
index 190fcb3da61..899903c1299 100644
--- a/backend/danswer/chat/prompts.yaml
+++ b/backend/danswer/chat/prompts.yaml
@@ -20,9 +20,15 @@ prompts:
# Task Prompt (as shown in UI)
task: >
If asked to find stuff give the exact find you got, if asked for links provide the links.
+
Make sure you pay attention to the timeframe of the data if specified in the query for example if asked for the latest ticket, it should be the earliest date you have in the knowledge base data set.
+
Do not use Knowledge base content older than 1 year in any response.
- Ifspecified for a certain data set only pull information from that data set such as Jira
+
+ If specified for a certain data set only pull information from that data set such as Jira.
+
+ If there are no relevant documents, refer to the chat history and your internal knowledge.
+
# Inject a statement at the end of system prompt to inform the LLM of the current date/time
# If the DANSWER_DATETIME_REPLACEMENT is set, the date/time is inserted there instead
# Format looks like: "October 16, 2023 14:30"
diff --git a/backend/danswer/configs/app_configs.py b/backend/danswer/configs/app_configs.py
index e33962971ce..577ef2f0aea 100644
--- a/backend/danswer/configs/app_configs.py
+++ b/backend/danswer/configs/app_configs.py
@@ -251,6 +251,7 @@
LOG_VESPA_TIMING_INFORMATION = (
os.environ.get("LOG_VESPA_TIMING_INFORMATION", "").lower() == "true"
)
+LOG_ENDPOINT_LATENCY = os.environ.get("LOG_ENDPOINT_LATENCY", "").lower() == "true"
# Anonymous usage telemetry
DISABLE_TELEMETRY = os.environ.get("DISABLE_TELEMETRY", "").lower() == "true"
diff --git a/backend/danswer/configs/constants.py b/backend/danswer/configs/constants.py
index b8acd3adff0..5f06a4acc07 100644
--- a/backend/danswer/configs/constants.py
+++ b/backend/danswer/configs/constants.py
@@ -93,6 +93,7 @@ class DocumentSource(str, Enum):
GOOGLE_SITES = "google_sites"
ZENDESK = "zendesk"
LOOPIO = "loopio"
+ DROPBOX = "dropbox"
SHAREPOINT = "sharepoint"
DISCOURSE = "discourse"
AXERO = "axero"
diff --git a/backend/danswer/configs/danswerbot_configs.py b/backend/danswer/configs/danswerbot_configs.py
index 1dc01bca1d2..5500d28c3ef 100644
--- a/backend/danswer/configs/danswerbot_configs.py
+++ b/backend/danswer/configs/danswerbot_configs.py
@@ -73,3 +73,7 @@
DANSWER_BOT_FEEDBACK_REMINDER = int(
os.environ.get("DANSWER_BOT_FEEDBACK_REMINDER") or 0
)
+# Set to True to rephrase the Slack users messages
+DANSWER_BOT_REPHRASE_MESSAGE = (
+ os.environ.get("DANSWER_BOT_REPHRASE_MESSAGE", "").lower() == "true"
+)
diff --git a/backend/danswer/connectors/dropbox/__init__.py b/backend/danswer/connectors/dropbox/__init__.py
new file mode 100644
index 00000000000..e69de29bb2d
diff --git a/backend/danswer/connectors/dropbox/connector.py b/backend/danswer/connectors/dropbox/connector.py
new file mode 100644
index 00000000000..2fd39948aaa
--- /dev/null
+++ b/backend/danswer/connectors/dropbox/connector.py
@@ -0,0 +1,151 @@
+from datetime import timezone
+from io import BytesIO
+from typing import Any
+
+from dropbox import Dropbox # type: ignore
+from dropbox.exceptions import ApiError # type:ignore
+from dropbox.files import FileMetadata # type:ignore
+from dropbox.files import FolderMetadata # type:ignore
+
+from danswer.configs.app_configs import INDEX_BATCH_SIZE
+from danswer.configs.constants import DocumentSource
+from danswer.connectors.interfaces import GenerateDocumentsOutput
+from danswer.connectors.interfaces import LoadConnector
+from danswer.connectors.interfaces import PollConnector
+from danswer.connectors.interfaces import SecondsSinceUnixEpoch
+from danswer.connectors.models import ConnectorMissingCredentialError
+from danswer.connectors.models import Document
+from danswer.connectors.models import Section
+from danswer.file_processing.extract_file_text import extract_file_text
+from danswer.utils.logger import setup_logger
+
+
+logger = setup_logger()
+
+
+class DropboxConnector(LoadConnector, PollConnector):
+ def __init__(self, batch_size: int = INDEX_BATCH_SIZE) -> None:
+ self.batch_size = batch_size
+ self.dropbox_client: Dropbox | None = None
+
+ def load_credentials(self, credentials: dict[str, Any]) -> dict[str, Any] | None:
+ self.dropbox_client = Dropbox(credentials["dropbox_access_token"])
+ return None
+
+ def _download_file(self, path: str) -> bytes:
+ """Download a single file from Dropbox."""
+ if self.dropbox_client is None:
+ raise ConnectorMissingCredentialError("Dropbox")
+ _, resp = self.dropbox_client.files_download(path)
+ return resp.content
+
+ def _get_shared_link(self, path: str) -> str:
+ """Create a shared link for a file in Dropbox."""
+ if self.dropbox_client is None:
+ raise ConnectorMissingCredentialError("Dropbox")
+
+ try:
+ # Check if a shared link already exists
+ shared_links = self.dropbox_client.sharing_list_shared_links(path=path)
+ if shared_links.links:
+ return shared_links.links[0].url
+
+ link_metadata = (
+ self.dropbox_client.sharing_create_shared_link_with_settings(path)
+ )
+ return link_metadata.url
+ except ApiError as err:
+ logger.exception(f"Failed to create a shared link for {path}: {err}")
+ return ""
+
+ def _yield_files_recursive(
+ self,
+ path: str,
+ start: SecondsSinceUnixEpoch | None,
+ end: SecondsSinceUnixEpoch | None,
+ ) -> GenerateDocumentsOutput:
+ """Yield files in batches from a specified Dropbox folder, including subfolders."""
+ if self.dropbox_client is None:
+ raise ConnectorMissingCredentialError("Dropbox")
+
+ result = self.dropbox_client.files_list_folder(
+ path,
+ limit=self.batch_size,
+ recursive=False,
+ include_non_downloadable_files=False,
+ )
+
+ while True:
+ batch: list[Document] = []
+ for entry in result.entries:
+ if isinstance(entry, FileMetadata):
+ modified_time = entry.client_modified
+ if modified_time.tzinfo is None:
+ # If no timezone info, assume it is UTC
+ modified_time = modified_time.replace(tzinfo=timezone.utc)
+ else:
+ # If not in UTC, translate it
+ modified_time = modified_time.astimezone(timezone.utc)
+
+ time_as_seconds = int(modified_time.timestamp())
+ if start and time_as_seconds < start:
+ continue
+ if end and time_as_seconds > end:
+ continue
+
+ downloaded_file = self._download_file(entry.path_display)
+ link = self._get_shared_link(entry.path_display)
+ try:
+ text = extract_file_text(entry.name, BytesIO(downloaded_file))
+ batch.append(
+ Document(
+ id=f"doc:{entry.id}",
+ sections=[Section(link=link, text=text)],
+ source=DocumentSource.DROPBOX,
+ semantic_identifier=entry.name,
+ doc_updated_at=modified_time,
+ metadata={"type": "article"},
+ )
+ )
+ except Exception as e:
+ logger.exception(
+ f"Error decoding file {entry.path_display} as utf-8 error occurred: {e}"
+ )
+
+ elif isinstance(entry, FolderMetadata):
+ yield from self._yield_files_recursive(entry.path_lower, start, end)
+
+ if batch:
+ yield batch
+
+ if not result.has_more:
+ break
+
+ result = self.dropbox_client.files_list_folder_continue(result.cursor)
+
+ def load_from_state(self) -> GenerateDocumentsOutput:
+ return self.poll_source(None, None)
+
+ def poll_source(
+ self, start: SecondsSinceUnixEpoch | None, end: SecondsSinceUnixEpoch | None
+ ) -> GenerateDocumentsOutput:
+ if self.dropbox_client is None:
+ raise ConnectorMissingCredentialError("Dropbox")
+
+ for batch in self._yield_files_recursive("", start, end):
+ yield batch
+
+ return None
+
+
+if __name__ == "__main__":
+ import os
+
+ connector = DropboxConnector()
+ connector.load_credentials(
+ {
+ "dropbox_access_token": os.environ["DROPBOX_ACCESS_TOKEN"],
+ }
+ )
+ document_batches = connector.load_from_state()
+ print(next(document_batches))
diff --git a/backend/danswer/connectors/factory.py b/backend/danswer/connectors/factory.py
index f68294be1ad..4873147e4f1 100644
--- a/backend/danswer/connectors/factory.py
+++ b/backend/danswer/connectors/factory.py
@@ -8,6 +8,7 @@
from danswer.connectors.danswer_jira.connector import JiraConnector
from danswer.connectors.discourse.connector import DiscourseConnector
from danswer.connectors.document360.connector import Document360Connector
+from danswer.connectors.dropbox.connector import DropboxConnector
from danswer.connectors.file.connector import LocalFileConnector
from danswer.connectors.github.connector import GithubConnector
from danswer.connectors.gitlab.connector import GitlabConnector
@@ -73,6 +74,7 @@ def identify_connector_class(
DocumentSource.GOOGLE_SITES: GoogleSitesConnector,
DocumentSource.ZENDESK: ZendeskConnector,
DocumentSource.LOOPIO: LoopioConnector,
+ DocumentSource.DROPBOX: DropboxConnector,
DocumentSource.SHAREPOINT: SharepointConnector,
DocumentSource.DISCOURSE: DiscourseConnector,
DocumentSource.AXERO: AxeroConnector,
diff --git a/backend/danswer/connectors/file/connector.py b/backend/danswer/connectors/file/connector.py
index 2e6a9081d88..431b23f272a 100644
--- a/backend/danswer/connectors/file/connector.py
+++ b/backend/danswer/connectors/file/connector.py
@@ -69,7 +69,7 @@ def _process_file(
if is_text_file_extension(file_name):
encoding = detect_encoding(file)
- file_content_raw, file_metadata = read_text_file(file, encoding=encoding)
+ file_content_raw, file_metadata = read_text_file(file, encoding=encoding, ignore_danswer_metadata=False)
# Using the PDF reader function directly to pass in password cleanly
elif extension == ".pdf":
diff --git a/backend/danswer/connectors/gitlab/connector.py b/backend/danswer/connectors/gitlab/connector.py
index 3b37dcdf5cf..55e358bbee0 100644
--- a/backend/danswer/connectors/gitlab/connector.py
+++ b/backend/danswer/connectors/gitlab/connector.py
@@ -6,6 +6,7 @@
from typing import Any
import gitlab
+import pytz
from danswer.configs.app_configs import INDEX_BATCH_SIZE
from danswer.configs.constants import DocumentSource
@@ -114,12 +115,14 @@ def _fetch_from_gitlab(
doc_batch: list[Document] = []
for mr in mr_batch:
mr.updated_at = datetime.strptime(
- mr.updated_at, "%Y-%m-%dT%H:%M:%S.%fZ"
+ mr.updated_at, "%Y-%m-%dT%H:%M:%S.%f%z"
)
- if start is not None and mr.updated_at < start:
+ if start is not None and mr.updated_at < start.replace(
+ tzinfo=pytz.UTC
+ ):
yield doc_batch
return
- if end is not None and mr.updated_at > end:
+ if end is not None and mr.updated_at > end.replace(tzinfo=pytz.UTC):
continue
doc_batch.append(_convert_merge_request_to_document(mr))
yield doc_batch
@@ -131,13 +134,17 @@ def _fetch_from_gitlab(
doc_batch = []
for issue in issue_batch:
issue.updated_at = datetime.strptime(
- issue.updated_at, "%Y-%m-%dT%H:%M:%S.%fZ"
+ issue.updated_at, "%Y-%m-%dT%H:%M:%S.%f%z"
)
- if start is not None and issue.updated_at < start:
- yield doc_batch
- return
- if end is not None and issue.updated_at > end:
- continue
+ if start is not None:
+ start = start.replace(tzinfo=pytz.UTC)
+ if issue.updated_at < start:
+ yield doc_batch
+ return
+ if end is not None:
+ end = end.replace(tzinfo=pytz.UTC)
+ if issue.updated_at > end:
+ continue
doc_batch.append(_convert_issue_to_document(issue))
yield doc_batch
diff --git a/backend/danswer/connectors/sharepoint/connector.py b/backend/danswer/connectors/sharepoint/connector.py
index d0e98f2fd61..0bd549f55e9 100644
--- a/backend/danswer/connectors/sharepoint/connector.py
+++ b/backend/danswer/connectors/sharepoint/connector.py
@@ -2,7 +2,8 @@
import os
from datetime import datetime
from datetime import timezone
-from typing import Any
+from typing import Any, Optional
+from dataclasses import dataclass, field
import msal # type: ignore
from office365.graph_client import GraphClient # type: ignore
@@ -58,6 +59,13 @@ def get_text_from_pptx_driveitem(driveitem_object: DriveItem) -> str:
file_content = driveitem_object.get_content().execute_query().value
return pptx_to_text(file=io.BytesIO(file_content))
+@dataclass
+class SiteData:
+ url: str
+ folder: Optional[str]
+ siteobjects: list = field(default_factory=list)
+ driveitems: list = field(default_factory=list)
+
class SharepointConnector(LoadConnector, PollConnector):
def __init__(
@@ -67,7 +75,19 @@ def __init__(
) -> None:
self.batch_size = batch_size
self.graph_client: GraphClient | None = None
- self.requested_site_list: list[str] = sites
+ self.site_data = self.extract_site_and_folder(sites)
+
+ @staticmethod
+ def extract_site_and_folder(site_urls: list[str]) -> list[SiteData]:
+ site_data_list = []
+ for url in site_urls:
+ parts = url.strip().split("/")
+ if "sites" in parts:
+ sites_index = parts.index("sites")
+ site_url = "/".join(parts[:sites_index + 2])
+ folder = parts[sites_index + 2] if len(parts) > sites_index + 2 else None
+ site_data_list.append(SiteData(url=site_url, folder=folder, siteobjects=[], driveitems=[]))
+ return site_data_list
def load_credentials(self, credentials: dict[str, Any]) -> dict[str, Any] | None:
aad_client_id = credentials["aad_client_id"]
@@ -94,7 +114,6 @@ def _acquire_token_func() -> dict[str, Any]:
def get_all_driveitem_objects(
self,
- site_object_list: list[Site],
start: datetime | None = None,
end: datetime | None = None,
) -> list[DriveItem]:
@@ -103,15 +122,24 @@ def get_all_driveitem_objects(
filter_str = f"last_modified_datetime ge {start.isoformat()} and last_modified_datetime le {end.isoformat()}"
driveitem_list = []
- for site_object in site_object_list:
- site_list_objects = site_object.lists.get().execute_query()
- for site_list_object in site_list_objects:
+ for element in self.site_data:
+ site_objects_list = []
+ for site_object in element.siteobjects:
+ site_objects_sublist = site_object.lists.get().execute_query()
+ site_objects_list.extend(site_objects_sublist)
+
+ for site_object in site_objects_list:
try:
- query = site_list_object.drive.root.get_files(True)
+ query = site_object.drive.root.get_files(True, 1000)
if filter_str:
query = query.filter(filter_str)
driveitems = query.execute_query()
- driveitem_list.extend(driveitems)
+ if element.folder:
+ filtered_driveitems = [item for item in driveitems if element.folder in item.parent_reference.path]
+ element.driveitems.extend(filtered_driveitems)
+ else:
+ element.driveitems.extend(driveitems)
+
except Exception:
# Sites include things that do not contain .drive.root so this fails
# but this is fine, as there are no actually documents in those
@@ -123,20 +151,12 @@ def get_all_site_objects(self) -> list[Site]:
if self.graph_client is None:
raise ConnectorMissingCredentialError("Sharepoint")
- site_object_list: list[Site] = []
-
- sites_object = self.graph_client.sites.get().execute_query()
-
- if len(self.requested_site_list) > 0:
- for requested_site in self.requested_site_list:
- adjusted_string = "/" + requested_site.replace(" ", "")
- for site_object in sites_object:
- if site_object.web_url.endswith(adjusted_string):
- site_object_list.append(site_object)
+ if self.site_data:
+ for element in self.site_data:
+ element.siteobjects = [self.graph_client.sites.get_by_url(element.url).get().execute_query()]
else:
- site_object_list.extend(sites_object)
-
- return site_object_list
+ site_objects = self.graph_client.sites.get().execute_query()
+ self.site_data = [SiteData(url=None, folder=None, siteobjects=site_objects, driveitems=[])]
def _fetch_from_sharepoint(
self, start: datetime | None = None, end: datetime | None = None
@@ -144,28 +164,24 @@ def _fetch_from_sharepoint(
if self.graph_client is None:
raise ConnectorMissingCredentialError("Sharepoint")
- site_object_list = self.get_all_site_objects()
-
- driveitem_list = self.get_all_driveitem_objects(
- site_object_list=site_object_list,
- start=start,
- end=end,
- )
+ self.get_all_site_objects()
+ self.get_all_driveitem_objects(start=start, end=end)
# goes over all urls, converts them into Document objects and then yields them in batches
doc_batch: list[Document] = []
batch_count = 0
- for driveitem_object in driveitem_list:
- logger.debug(f"Processing: {driveitem_object.web_url}")
- doc_batch.append(
- self.convert_driveitem_object_to_document(driveitem_object)
- )
+ for element in self.site_data:
+ for driveitem_object in element.driveitems:
+ logger.debug(f"Processing: {driveitem_object.web_url}")
+ doc_batch.append(
+ self.convert_driveitem_object_to_document(driveitem_object)
+ )
- batch_count += 1
- if batch_count >= self.batch_size:
- yield doc_batch
- batch_count = 0
- doc_batch = []
+ batch_count += 1
+ if batch_count >= self.batch_size:
+ yield doc_batch
+ batch_count = 0
+ doc_batch = []
yield doc_batch
def convert_driveitem_object_to_document(
@@ -230,4 +246,4 @@ def poll_source(
}
)
document_batches = connector.load_from_state()
- print(next(document_batches))
+ print(next(document_batches))
\ No newline at end of file
diff --git a/backend/danswer/danswerbot/slack/listener.py b/backend/danswer/danswerbot/slack/listener.py
index ce7c9eda216..2ccce9fa773 100644
--- a/backend/danswer/danswerbot/slack/listener.py
+++ b/backend/danswer/danswerbot/slack/listener.py
@@ -10,6 +10,7 @@
from sqlalchemy.orm import Session
from danswer.configs.constants import MessageType
+from danswer.configs.danswerbot_configs import DANSWER_BOT_REPHRASE_MESSAGE
from danswer.configs.danswerbot_configs import DANSWER_BOT_RESPOND_EVERY_CHANNEL
from danswer.configs.danswerbot_configs import NOTIFY_SLACKBOT_NO_ANSWER
from danswer.danswerbot.slack.config import get_slack_bot_config_for_channel
@@ -40,6 +41,7 @@
from danswer.danswerbot.slack.utils import get_danswer_bot_app_id
from danswer.danswerbot.slack.utils import read_slack_thread
from danswer.danswerbot.slack.utils import remove_danswer_bot_tag
+from danswer.danswerbot.slack.utils import rephrase_slack_message
from danswer.danswerbot.slack.utils import respond_in_thread
from danswer.db.embedding_model import get_current_db_embedding_model
from danswer.db.engine import get_sqlalchemy_engine
@@ -202,6 +204,14 @@ def build_request_details(
msg = remove_danswer_bot_tag(msg, client=client.web_client)
+ if DANSWER_BOT_REPHRASE_MESSAGE:
+ logger.info(f"Rephrasing Slack message. Original message: {msg}")
+ try:
+ msg = rephrase_slack_message(msg)
+ logger.info(f"Rephrased message: {msg}")
+ except Exception as e:
+ logger.error(f"Error while trying to rephrase the Slack message: {e}")
+
if tagged:
logger.info("User tagged DanswerBot")
diff --git a/backend/danswer/danswerbot/slack/utils.py b/backend/danswer/danswerbot/slack/utils.py
index 5895dc52f91..892734e9995 100644
--- a/backend/danswer/danswerbot/slack/utils.py
+++ b/backend/danswer/danswerbot/slack/utils.py
@@ -29,7 +29,12 @@
from danswer.danswerbot.slack.tokens import fetch_tokens
from danswer.db.engine import get_sqlalchemy_engine
from danswer.db.users import get_user_by_email
+from danswer.llm.exceptions import GenAIDisabledException
+from danswer.llm.factory import get_default_llm
+from danswer.llm.utils import dict_based_prompt_to_langchain_prompt
+from danswer.llm.utils import message_to_string
from danswer.one_shot_answer.models import ThreadMessage
+from danswer.prompts.miscellaneous_prompts import SLACK_LANGUAGE_REPHRASE_PROMPT
from danswer.utils.logger import setup_logger
from danswer.utils.telemetry import optional_telemetry
from danswer.utils.telemetry import RecordType
@@ -41,6 +46,30 @@
DANSWER_BOT_APP_ID: str | None = None
+def rephrase_slack_message(msg: str) -> str:
+ def _get_rephrase_message() -> list[dict[str, str]]:
+ messages = [
+ {
+ "role": "user",
+ "content": SLACK_LANGUAGE_REPHRASE_PROMPT.format(query=msg),
+ },
+ ]
+
+ return messages
+
+ try:
+ llm = get_default_llm(use_fast_llm=False, timeout=5)
+ except GenAIDisabledException:
+ logger.warning("Unable to rephrase Slack user message, Gen AI disabled")
+ return msg
+ messages = _get_rephrase_message()
+ filled_llm_prompt = dict_based_prompt_to_langchain_prompt(messages)
+ model_output = message_to_string(llm.invoke(filled_llm_prompt))
+ logger.debug(model_output)
+
+ return model_output
+
+
def update_emote_react(
emoji: str,
channel: str,
diff --git a/backend/danswer/db/chat.py b/backend/danswer/db/chat.py
index 3e006b021b2..50458efd8d8 100644
--- a/backend/danswer/db/chat.py
+++ b/backend/danswer/db/chat.py
@@ -1,40 +1,25 @@
-from collections.abc import Sequence
-from functools import lru_cache
from uuid import UUID
from sqlalchemy import delete
-from sqlalchemy import func
-from sqlalchemy import not_
from sqlalchemy import nullsfirst
from sqlalchemy import or_
from sqlalchemy import select
-from sqlalchemy import update
from sqlalchemy.exc import MultipleResultsFound
from sqlalchemy.orm import Session
from danswer.auth.schemas import UserRole
from danswer.configs.chat_configs import HARD_DELETE_CHATS
from danswer.configs.constants import MessageType
-from danswer.db.constants import SLACK_BOT_PERSONA_PREFIX
-from danswer.db.engine import get_sqlalchemy_engine
from danswer.db.models import ChatMessage
from danswer.db.models import ChatSession
from danswer.db.models import ChatSessionSharedStatus
-from danswer.db.models import DocumentSet as DBDocumentSet
-from danswer.db.models import Persona
-from danswer.db.models import Persona__User
-from danswer.db.models import Persona__UserGroup
from danswer.db.models import Prompt
from danswer.db.models import SearchDoc
from danswer.db.models import SearchDoc as DBSearchDoc
-from danswer.db.models import StarterMessage
-from danswer.db.models import Tool
from danswer.db.models import User
-from danswer.db.models import User__UserGroup
from danswer.file_store.models import FileDescriptor
from danswer.llm.override_models import LLMOverride
from danswer.llm.override_models import PromptOverride
-from danswer.search.enums import RecencyBiasSetting
from danswer.search.models import RetrievalDocs
from danswer.search.models import SavedSearchDoc
from danswer.search.models import SearchDoc as ServerSearchDoc
@@ -357,396 +342,6 @@ def get_prompt_by_id(
return prompt
-@lru_cache()
-def get_default_prompt() -> Prompt:
- with Session(get_sqlalchemy_engine()) as db_session:
- stmt = select(Prompt).where(Prompt.id == 0)
-
- result = db_session.execute(stmt)
- prompt = result.scalar_one_or_none()
-
- if prompt is None:
- raise RuntimeError("Default Prompt not found")
-
- return prompt
-
-
-def get_persona_by_id(
- persona_id: int,
- # if user is `None` assume the user is an admin or auth is disabled
- user: User | None,
- db_session: Session,
- include_deleted: bool = False,
-) -> Persona:
- stmt = select(Persona).where(Persona.id == persona_id)
-
- # if user is an admin, they should have access to all Personas
- if user is not None and user.role != UserRole.ADMIN:
- stmt = stmt.where(or_(Persona.user_id == user.id, Persona.user_id.is_(None)))
-
- if not include_deleted:
- stmt = stmt.where(Persona.deleted.is_(False))
-
- result = db_session.execute(stmt)
- persona = result.scalar_one_or_none()
-
- if persona is None:
- raise ValueError(
- f"Persona with ID {persona_id} does not exist or does not belong to user"
- )
-
- return persona
-
-
-def get_prompts_by_ids(prompt_ids: list[int], db_session: Session) -> Sequence[Prompt]:
- """Unsafe, can fetch prompts from all users"""
- if not prompt_ids:
- return []
- prompts = db_session.scalars(select(Prompt).where(Prompt.id.in_(prompt_ids))).all()
-
- return prompts
-
-
-def get_personas_by_ids(
- persona_ids: list[int], db_session: Session
-) -> Sequence[Persona]:
- """Unsafe, can fetch personas from all users"""
- if not persona_ids:
- return []
- personas = db_session.scalars(
- select(Persona).where(Persona.id.in_(persona_ids))
- ).all()
-
- return personas
-
-
-def get_prompt_by_name(
- prompt_name: str, user: User | None, db_session: Session
-) -> Prompt | None:
- stmt = select(Prompt).where(Prompt.name == prompt_name)
-
- # if user is not specified OR they are an admin, they should
- # have access to all prompts, so this where clause is not needed
- if user and user.role != UserRole.ADMIN:
- stmt = stmt.where(Prompt.user_id == user.id)
-
- result = db_session.execute(stmt).scalar_one_or_none()
- return result
-
-
-def get_persona_by_name(
- persona_name: str, user: User | None, db_session: Session
-) -> Persona | None:
- """Admins can see all, regular users can only fetch their own.
- If user is None, assume the user is an admin or auth is disabled."""
- stmt = select(Persona).where(Persona.name == persona_name)
- if user and user.role != UserRole.ADMIN:
- stmt = stmt.where(Persona.user_id == user.id)
- result = db_session.execute(stmt).scalar_one_or_none()
- return result
-
-
-def upsert_prompt(
- user: User | None,
- name: str,
- description: str,
- system_prompt: str,
- task_prompt: str,
- include_citations: bool,
- datetime_aware: bool,
- personas: list[Persona] | None,
- db_session: Session,
- prompt_id: int | None = None,
- default_prompt: bool = True,
- commit: bool = True,
-) -> Prompt:
- if prompt_id is not None:
- prompt = db_session.query(Prompt).filter_by(id=prompt_id).first()
- else:
- prompt = get_prompt_by_name(prompt_name=name, user=user, db_session=db_session)
-
- if prompt:
- if not default_prompt and prompt.default_prompt:
- raise ValueError("Cannot update default prompt with non-default.")
-
- prompt.name = name
- prompt.description = description
- prompt.system_prompt = system_prompt
- prompt.task_prompt = task_prompt
- prompt.include_citations = include_citations
- prompt.datetime_aware = datetime_aware
- prompt.default_prompt = default_prompt
-
- if personas is not None:
- prompt.personas.clear()
- prompt.personas = personas
-
- else:
- prompt = Prompt(
- id=prompt_id,
- user_id=user.id if user else None,
- name=name,
- description=description,
- system_prompt=system_prompt,
- task_prompt=task_prompt,
- include_citations=include_citations,
- datetime_aware=datetime_aware,
- default_prompt=default_prompt,
- personas=personas or [],
- )
- db_session.add(prompt)
-
- if commit:
- db_session.commit()
- else:
- # Flush the session so that the Prompt has an ID
- db_session.flush()
-
- return prompt
-
-
-def upsert_persona(
- user: User | None,
- name: str,
- description: str,
- num_chunks: float,
- llm_relevance_filter: bool,
- llm_filter_extraction: bool,
- recency_bias: RecencyBiasSetting,
- prompts: list[Prompt] | None,
- document_sets: list[DBDocumentSet] | None,
- llm_model_provider_override: str | None,
- llm_model_version_override: str | None,
- starter_messages: list[StarterMessage] | None,
- is_public: bool,
- db_session: Session,
- tool_ids: list[int] | None = None,
- persona_id: int | None = None,
- default_persona: bool = False,
- commit: bool = True,
-) -> Persona:
- if persona_id is not None:
- persona = db_session.query(Persona).filter_by(id=persona_id).first()
- else:
- persona = get_persona_by_name(
- persona_name=name, user=user, db_session=db_session
- )
-
- # Fetch and attach tools by IDs
- tools = None
- if tool_ids is not None:
- tools = db_session.query(Tool).filter(Tool.id.in_(tool_ids)).all()
- if not tools and tool_ids:
- raise ValueError("Tools not found")
-
- if persona:
- if not default_persona and persona.default_persona:
- raise ValueError("Cannot update default persona with non-default.")
-
- persona.name = name
- persona.description = description
- persona.num_chunks = num_chunks
- persona.llm_relevance_filter = llm_relevance_filter
- persona.llm_filter_extraction = llm_filter_extraction
- persona.recency_bias = recency_bias
- persona.default_persona = default_persona
- persona.llm_model_provider_override = llm_model_provider_override
- persona.llm_model_version_override = llm_model_version_override
- persona.starter_messages = starter_messages
- persona.deleted = False # Un-delete if previously deleted
- persona.is_public = is_public
-
- # Do not delete any associations manually added unless
- # a new updated list is provided
- if document_sets is not None:
- persona.document_sets.clear()
- persona.document_sets = document_sets or []
-
- if prompts is not None:
- persona.prompts.clear()
- persona.prompts = prompts
-
- if tools is not None:
- persona.tools = tools
-
- else:
- persona = Persona(
- id=persona_id,
- user_id=user.id if user else None,
- is_public=is_public,
- name=name,
- description=description,
- num_chunks=num_chunks,
- llm_relevance_filter=llm_relevance_filter,
- llm_filter_extraction=llm_filter_extraction,
- recency_bias=recency_bias,
- default_persona=default_persona,
- prompts=prompts or [],
- document_sets=document_sets or [],
- llm_model_provider_override=llm_model_provider_override,
- llm_model_version_override=llm_model_version_override,
- starter_messages=starter_messages,
- tools=tools or [],
- )
- db_session.add(persona)
-
- if commit:
- db_session.commit()
- else:
- # flush the session so that the persona has an ID
- db_session.flush()
-
- return persona
-
-
-def mark_prompt_as_deleted(
- prompt_id: int,
- user: User | None,
- db_session: Session,
-) -> None:
- prompt = get_prompt_by_id(prompt_id=prompt_id, user=user, db_session=db_session)
- prompt.deleted = True
- db_session.commit()
-
-
-def mark_persona_as_deleted(
- persona_id: int,
- user: User | None,
- db_session: Session,
-) -> None:
- persona = get_persona_by_id(persona_id=persona_id, user=user, db_session=db_session)
- persona.deleted = True
- db_session.commit()
-
-
-def mark_persona_as_not_deleted(
- persona_id: int,
- user: User | None,
- db_session: Session,
-) -> None:
- persona = get_persona_by_id(
- persona_id=persona_id, user=user, db_session=db_session, include_deleted=True
- )
- if persona.deleted:
- persona.deleted = False
- db_session.commit()
- else:
- raise ValueError(f"Persona with ID {persona_id} is not deleted.")
-
-
-def mark_delete_persona_by_name(
- persona_name: str, db_session: Session, is_default: bool = True
-) -> None:
- stmt = (
- update(Persona)
- .where(Persona.name == persona_name, Persona.default_persona == is_default)
- .values(deleted=True)
- )
-
- db_session.execute(stmt)
- db_session.commit()
-
-
-def delete_old_default_personas(
- db_session: Session,
-) -> None:
- """Note, this locks out the Summarize and Paraphrase personas for now
- Need a more graceful fix later or those need to never have IDs"""
- stmt = (
- update(Persona)
- .where(Persona.default_persona, Persona.id > 0)
- .values(deleted=True, name=func.concat(Persona.name, "_old"))
- )
-
- db_session.execute(stmt)
- db_session.commit()
-
-
-def update_persona_visibility(
- persona_id: int,
- is_visible: bool,
- db_session: Session,
-) -> None:
- persona = get_persona_by_id(persona_id=persona_id, user=None, db_session=db_session)
- persona.is_visible = is_visible
- db_session.commit()
-
-
-def update_all_personas_display_priority(
- display_priority_map: dict[int, int],
- db_session: Session,
-) -> None:
- """Updates the display priority of all lives Personas"""
- personas = get_personas(user_id=None, db_session=db_session)
- available_persona_ids = {persona.id for persona in personas}
- if available_persona_ids != set(display_priority_map.keys()):
- raise ValueError("Invalid persona IDs provided")
-
- for persona in personas:
- persona.display_priority = display_priority_map[persona.id]
-
- db_session.commit()
-
-
-def get_prompts(
- user_id: UUID | None,
- db_session: Session,
- include_default: bool = True,
- include_deleted: bool = False,
-) -> Sequence[Prompt]:
- stmt = select(Prompt).where(
- or_(Prompt.user_id == user_id, Prompt.user_id.is_(None))
- )
-
- if not include_default:
- stmt = stmt.where(Prompt.default_prompt.is_(False))
- if not include_deleted:
- stmt = stmt.where(Prompt.deleted.is_(False))
-
- return db_session.scalars(stmt).all()
-
-
-def get_personas(
- # if user_id is `None` assume the user is an admin or auth is disabled
- user_id: UUID | None,
- db_session: Session,
- include_default: bool = True,
- include_slack_bot_personas: bool = False,
- include_deleted: bool = False,
-) -> Sequence[Persona]:
- stmt = select(Persona).distinct()
- if user_id is not None:
- # Subquery to find all groups the user belongs to
- user_groups_subquery = (
- select(User__UserGroup.user_group_id)
- .where(User__UserGroup.user_id == user_id)
- .subquery()
- )
-
- # Include personas where the user is directly related or part of a user group that has access
- access_conditions = or_(
- Persona.is_public == True, # noqa: E712
- Persona.id.in_( # User has access through list of users with access
- select(Persona__User.persona_id).where(Persona__User.user_id == user_id)
- ),
- Persona.id.in_( # User is part of a group that has access
- select(Persona__UserGroup.persona_id).where(
- Persona__UserGroup.user_group_id.in_(user_groups_subquery) # type: ignore
- )
- ),
- )
- stmt = stmt.where(access_conditions)
-
- if not include_default:
- stmt = stmt.where(Persona.default_persona.is_(False))
- if not include_slack_bot_personas:
- stmt = stmt.where(not_(Persona.name.startswith(SLACK_BOT_PERSONA_PREFIX)))
- if not include_deleted:
- stmt = stmt.where(Persona.deleted.is_(False))
-
- return db_session.scalars(stmt).all()
-
-
def get_doc_query_identifiers_from_model(
search_doc_ids: list[int],
chat_session: ChatSession,
@@ -867,15 +462,3 @@ def translate_db_message_to_chat_message_detail(
)
return chat_msg_detail
-
-
-def delete_persona_by_name(
- persona_name: str, db_session: Session, is_default: bool = True
-) -> None:
- stmt = delete(Persona).where(
- Persona.name == persona_name, Persona.default_persona == is_default
- )
-
- db_session.execute(stmt)
-
- db_session.commit()
diff --git a/backend/danswer/db/engine.py b/backend/danswer/db/engine.py
index a8938ca0155..b8a7f858fd2 100644
--- a/backend/danswer/db/engine.py
+++ b/backend/danswer/db/engine.py
@@ -61,7 +61,9 @@ def get_sqlalchemy_engine() -> Engine:
global _SYNC_ENGINE
if _SYNC_ENGINE is None:
connection_string = build_connection_string(db_api=SYNC_DB_API)
- _SYNC_ENGINE = create_engine(connection_string, connect_args=connect_args)
+ _SYNC_ENGINE = create_engine(
+ connection_string, pool_size=40, max_overflow=10, connect_args=connect_args
+ )
return _SYNC_ENGINE
@@ -71,7 +73,7 @@ def get_sqlalchemy_async_engine() -> AsyncEngine:
if _ASYNC_ENGINE is None:
connection_string = build_connection_string()
_ASYNC_ENGINE = create_async_engine(
- connection_string, connect_args=connect_args
+ connection_string, pool_size=40, max_overflow=10, connect_args=connect_args
)
return _ASYNC_ENGINE
@@ -95,4 +97,27 @@ async def get_async_session() -> AsyncGenerator[AsyncSession, None]:
yield async_session
+async def warm_up_connections(
+ sync_connections_to_warm_up: int = 10, async_connections_to_warm_up: int = 10
+) -> None:
+ sync_postgres_engine = get_sqlalchemy_engine()
+ connections = [
+ sync_postgres_engine.connect() for _ in range(sync_connections_to_warm_up)
+ ]
+ for conn in connections:
+ conn.execute(text("SELECT 1"))
+ for conn in connections:
+ conn.close()
+
+ async_postgres_engine = get_sqlalchemy_async_engine()
+ async_connections = [
+ await async_postgres_engine.connect()
+ for _ in range(async_connections_to_warm_up)
+ ]
+ for async_conn in async_connections:
+ await async_conn.execute(text("SELECT 1"))
+ for async_conn in async_connections:
+ await async_conn.close()
+
+
SessionFactory = sessionmaker(bind=get_sqlalchemy_engine())
diff --git a/backend/danswer/db/models.py b/backend/danswer/db/models.py
index a77223394f7..3b4b67f0947 100644
--- a/backend/danswer/db/models.py
+++ b/backend/danswer/db/models.py
@@ -108,6 +108,19 @@ class User(SQLAlchemyBaseUserTableUUID, Base):
role: Mapped[UserRole] = mapped_column(
Enum(UserRole, native_enum=False, default=UserRole.BASIC)
)
+
+ """
+ Preferences probably should be in a separate table at some point, but for now
+ putting here for simpicity
+ """
+
+ # if specified, controls the assistants that are shown to the user + their order
+ # if not specified, all assistants are shown
+ chosen_assistants: Mapped[list[int]] = mapped_column(
+ postgresql.ARRAY(Integer), nullable=True
+ )
+
+ # relationships
credentials: Mapped[list["Credential"]] = relationship(
"Credential", back_populates="user", lazy="joined"
)
diff --git a/backend/danswer/db/persona.py b/backend/danswer/db/persona.py
index 6fa65135dca..88a72cc9331 100644
--- a/backend/danswer/db/persona.py
+++ b/backend/danswer/db/persona.py
@@ -1,15 +1,30 @@
+from collections.abc import Sequence
+from functools import lru_cache
from uuid import UUID
from fastapi import HTTPException
+from sqlalchemy import delete
+from sqlalchemy import func
+from sqlalchemy import not_
+from sqlalchemy import or_
from sqlalchemy import select
+from sqlalchemy import update
from sqlalchemy.orm import Session
-from danswer.db.chat import get_prompts_by_ids
-from danswer.db.chat import upsert_persona
+from danswer.auth.schemas import UserRole
+from danswer.db.constants import SLACK_BOT_PERSONA_PREFIX
from danswer.db.document_set import get_document_sets_by_ids
+from danswer.db.engine import get_sqlalchemy_engine
+from danswer.db.models import DocumentSet
from danswer.db.models import Persona
from danswer.db.models import Persona__User
+from danswer.db.models import Persona__UserGroup
+from danswer.db.models import Prompt
+from danswer.db.models import StarterMessage
+from danswer.db.models import Tool
from danswer.db.models import User
+from danswer.db.models import User__UserGroup
+from danswer.search.enums import RecencyBiasSetting
from danswer.server.features.persona.models import CreatePersonaRequest
from danswer.server.features.persona.models import PersonaSnapshot
from danswer.utils.logger import setup_logger
@@ -45,6 +60,7 @@ def create_update_persona(
user: User | None,
db_session: Session,
) -> PersonaSnapshot:
+ """Higher level function than upsert_persona, although either is valid to use."""
# Permission to actually use these is checked later
document_sets = list(
get_document_sets_by_ids(
@@ -97,5 +113,486 @@ def create_update_persona(
return PersonaSnapshot.from_model(persona)
+def update_persona_shared_users(
+ persona_id: int,
+ user_ids: list[UUID],
+ user: User | None,
+ db_session: Session,
+) -> None:
+ """Simplified version of `create_update_persona` which only touches the
+ accessibility rather than any of the logic (e.g. prompt, connected data sources,
+ etc.)."""
+ persona = fetch_persona_by_id(db_session=db_session, persona_id=persona_id)
+ if not persona:
+ raise HTTPException(
+ status_code=404, detail=f"Persona with ID {persona_id} not found"
+ )
+
+ check_user_can_edit_persona(user=user, persona=persona)
+
+ if persona.is_public:
+ raise HTTPException(status_code=400, detail="Cannot share public persona")
+
+ versioned_make_persona_private = fetch_versioned_implementation(
+ "danswer.db.persona", "make_persona_private"
+ )
+
+ # Privatize Persona
+ versioned_make_persona_private(
+ persona_id=persona_id,
+ user_ids=user_ids,
+ group_ids=None,
+ db_session=db_session,
+ )
+
+
def fetch_persona_by_id(db_session: Session, persona_id: int) -> Persona | None:
return db_session.scalar(select(Persona).where(Persona.id == persona_id))
+
+
+def get_prompts(
+ user_id: UUID | None,
+ db_session: Session,
+ include_default: bool = True,
+ include_deleted: bool = False,
+) -> Sequence[Prompt]:
+ stmt = select(Prompt).where(
+ or_(Prompt.user_id == user_id, Prompt.user_id.is_(None))
+ )
+
+ if not include_default:
+ stmt = stmt.where(Prompt.default_prompt.is_(False))
+ if not include_deleted:
+ stmt = stmt.where(Prompt.deleted.is_(False))
+
+ return db_session.scalars(stmt).all()
+
+
+def get_personas(
+ # if user_id is `None` assume the user is an admin or auth is disabled
+ user_id: UUID | None,
+ db_session: Session,
+ include_default: bool = True,
+ include_slack_bot_personas: bool = False,
+ include_deleted: bool = False,
+) -> Sequence[Persona]:
+ stmt = select(Persona).distinct()
+ if user_id is not None:
+ # Subquery to find all groups the user belongs to
+ user_groups_subquery = (
+ select(User__UserGroup.user_group_id)
+ .where(User__UserGroup.user_id == user_id)
+ .subquery()
+ )
+
+ # Include personas where the user is directly related or part of a user group that has access
+ access_conditions = or_(
+ Persona.is_public == True, # noqa: E712
+ Persona.id.in_( # User has access through list of users with access
+ select(Persona__User.persona_id).where(Persona__User.user_id == user_id)
+ ),
+ Persona.id.in_( # User is part of a group that has access
+ select(Persona__UserGroup.persona_id).where(
+ Persona__UserGroup.user_group_id.in_(user_groups_subquery) # type: ignore
+ )
+ ),
+ )
+ stmt = stmt.where(access_conditions)
+
+ if not include_default:
+ stmt = stmt.where(Persona.default_persona.is_(False))
+ if not include_slack_bot_personas:
+ stmt = stmt.where(not_(Persona.name.startswith(SLACK_BOT_PERSONA_PREFIX)))
+ if not include_deleted:
+ stmt = stmt.where(Persona.deleted.is_(False))
+
+ return db_session.scalars(stmt).all()
+
+
+def mark_persona_as_deleted(
+ persona_id: int,
+ user: User | None,
+ db_session: Session,
+) -> None:
+ persona = get_persona_by_id(persona_id=persona_id, user=user, db_session=db_session)
+ persona.deleted = True
+ db_session.commit()
+
+
+def mark_persona_as_not_deleted(
+ persona_id: int,
+ user: User | None,
+ db_session: Session,
+) -> None:
+ persona = get_persona_by_id(
+ persona_id=persona_id, user=user, db_session=db_session, include_deleted=True
+ )
+ if persona.deleted:
+ persona.deleted = False
+ db_session.commit()
+ else:
+ raise ValueError(f"Persona with ID {persona_id} is not deleted.")
+
+
+def mark_delete_persona_by_name(
+ persona_name: str, db_session: Session, is_default: bool = True
+) -> None:
+ stmt = (
+ update(Persona)
+ .where(Persona.name == persona_name, Persona.default_persona == is_default)
+ .values(deleted=True)
+ )
+
+ db_session.execute(stmt)
+ db_session.commit()
+
+
+def update_all_personas_display_priority(
+ display_priority_map: dict[int, int],
+ db_session: Session,
+) -> None:
+ """Updates the display priority of all lives Personas"""
+ personas = get_personas(user_id=None, db_session=db_session)
+ available_persona_ids = {persona.id for persona in personas}
+ if available_persona_ids != set(display_priority_map.keys()):
+ raise ValueError("Invalid persona IDs provided")
+
+ for persona in personas:
+ persona.display_priority = display_priority_map[persona.id]
+
+ db_session.commit()
+
+
+def upsert_prompt(
+ user: User | None,
+ name: str,
+ description: str,
+ system_prompt: str,
+ task_prompt: str,
+ include_citations: bool,
+ datetime_aware: bool,
+ personas: list[Persona] | None,
+ db_session: Session,
+ prompt_id: int | None = None,
+ default_prompt: bool = True,
+ commit: bool = True,
+) -> Prompt:
+ if prompt_id is not None:
+ prompt = db_session.query(Prompt).filter_by(id=prompt_id).first()
+ else:
+ prompt = get_prompt_by_name(prompt_name=name, user=user, db_session=db_session)
+
+ if prompt:
+ if not default_prompt and prompt.default_prompt:
+ raise ValueError("Cannot update default prompt with non-default.")
+
+ prompt.name = name
+ prompt.description = description
+ prompt.system_prompt = system_prompt
+ prompt.task_prompt = task_prompt
+ prompt.include_citations = include_citations
+ prompt.datetime_aware = datetime_aware
+ prompt.default_prompt = default_prompt
+
+ if personas is not None:
+ prompt.personas.clear()
+ prompt.personas = personas
+
+ else:
+ prompt = Prompt(
+ id=prompt_id,
+ user_id=user.id if user else None,
+ name=name,
+ description=description,
+ system_prompt=system_prompt,
+ task_prompt=task_prompt,
+ include_citations=include_citations,
+ datetime_aware=datetime_aware,
+ default_prompt=default_prompt,
+ personas=personas or [],
+ )
+ db_session.add(prompt)
+
+ if commit:
+ db_session.commit()
+ else:
+ # Flush the session so that the Prompt has an ID
+ db_session.flush()
+
+ return prompt
+
+
+def upsert_persona(
+ user: User | None,
+ name: str,
+ description: str,
+ num_chunks: float,
+ llm_relevance_filter: bool,
+ llm_filter_extraction: bool,
+ recency_bias: RecencyBiasSetting,
+ prompts: list[Prompt] | None,
+ document_sets: list[DocumentSet] | None,
+ llm_model_provider_override: str | None,
+ llm_model_version_override: str | None,
+ starter_messages: list[StarterMessage] | None,
+ is_public: bool,
+ db_session: Session,
+ tool_ids: list[int] | None = None,
+ persona_id: int | None = None,
+ default_persona: bool = False,
+ commit: bool = True,
+) -> Persona:
+ if persona_id is not None:
+ persona = db_session.query(Persona).filter_by(id=persona_id).first()
+ else:
+ persona = get_persona_by_name(
+ persona_name=name, user=user, db_session=db_session
+ )
+
+ # Fetch and attach tools by IDs
+ tools = None
+ if tool_ids is not None:
+ tools = db_session.query(Tool).filter(Tool.id.in_(tool_ids)).all()
+ if not tools and tool_ids:
+ raise ValueError("Tools not found")
+
+ if persona:
+ if not default_persona and persona.default_persona:
+ raise ValueError("Cannot update default persona with non-default.")
+
+ check_user_can_edit_persona(user=user, persona=persona)
+
+ persona.name = name
+ persona.description = description
+ persona.num_chunks = num_chunks
+ persona.llm_relevance_filter = llm_relevance_filter
+ persona.llm_filter_extraction = llm_filter_extraction
+ persona.recency_bias = recency_bias
+ persona.default_persona = default_persona
+ persona.llm_model_provider_override = llm_model_provider_override
+ persona.llm_model_version_override = llm_model_version_override
+ persona.starter_messages = starter_messages
+ persona.deleted = False # Un-delete if previously deleted
+ persona.is_public = is_public
+
+ # Do not delete any associations manually added unless
+ # a new updated list is provided
+ if document_sets is not None:
+ persona.document_sets.clear()
+ persona.document_sets = document_sets or []
+
+ if prompts is not None:
+ persona.prompts.clear()
+ persona.prompts = prompts
+
+ if tools is not None:
+ persona.tools = tools
+
+ else:
+ persona = Persona(
+ id=persona_id,
+ user_id=user.id if user else None,
+ is_public=is_public,
+ name=name,
+ description=description,
+ num_chunks=num_chunks,
+ llm_relevance_filter=llm_relevance_filter,
+ llm_filter_extraction=llm_filter_extraction,
+ recency_bias=recency_bias,
+ default_persona=default_persona,
+ prompts=prompts or [],
+ document_sets=document_sets or [],
+ llm_model_provider_override=llm_model_provider_override,
+ llm_model_version_override=llm_model_version_override,
+ starter_messages=starter_messages,
+ tools=tools or [],
+ )
+ db_session.add(persona)
+
+ if commit:
+ db_session.commit()
+ else:
+ # flush the session so that the persona has an ID
+ db_session.flush()
+
+ return persona
+
+
+def mark_prompt_as_deleted(
+ prompt_id: int,
+ user: User | None,
+ db_session: Session,
+) -> None:
+ prompt = get_prompt_by_id(prompt_id=prompt_id, user=user, db_session=db_session)
+ prompt.deleted = True
+ db_session.commit()
+
+
+def delete_old_default_personas(
+ db_session: Session,
+) -> None:
+ """Note, this locks out the Summarize and Paraphrase personas for now
+ Need a more graceful fix later or those need to never have IDs"""
+ stmt = (
+ update(Persona)
+ .where(Persona.default_persona, Persona.id > 0)
+ .values(deleted=True, name=func.concat(Persona.name, "_old"))
+ )
+
+ db_session.execute(stmt)
+ db_session.commit()
+
+
+def update_persona_visibility(
+ persona_id: int,
+ is_visible: bool,
+ db_session: Session,
+) -> None:
+ persona = get_persona_by_id(persona_id=persona_id, user=None, db_session=db_session)
+ persona.is_visible = is_visible
+ db_session.commit()
+
+
+def check_user_can_edit_persona(user: User | None, persona: Persona) -> None:
+ # if user is None, assume that no-auth is turned on
+ if user is None:
+ return
+
+ # admins can edit everything
+ if user.role == UserRole.ADMIN:
+ return
+
+ # otherwise, make sure user owns persona
+ if persona.user_id != user.id:
+ raise HTTPException(
+ status_code=403,
+ detail=f"User not authorized to edit persona with ID {persona.id}",
+ )
+
+
+def get_prompts_by_ids(prompt_ids: list[int], db_session: Session) -> Sequence[Prompt]:
+ """Unsafe, can fetch prompts from all users"""
+ if not prompt_ids:
+ return []
+ prompts = db_session.scalars(select(Prompt).where(Prompt.id.in_(prompt_ids))).all()
+
+ return prompts
+
+
+def get_prompt_by_id(
+ prompt_id: int,
+ user: User | None,
+ db_session: Session,
+ include_deleted: bool = False,
+) -> Prompt:
+ stmt = select(Prompt).where(Prompt.id == prompt_id)
+
+ # if user is not specified OR they are an admin, they should
+ # have access to all prompts, so this where clause is not needed
+ if user and user.role != UserRole.ADMIN:
+ stmt = stmt.where(or_(Prompt.user_id == user.id, Prompt.user_id.is_(None)))
+
+ if not include_deleted:
+ stmt = stmt.where(Prompt.deleted.is_(False))
+
+ result = db_session.execute(stmt)
+ prompt = result.scalar_one_or_none()
+
+ if prompt is None:
+ raise ValueError(
+ f"Prompt with ID {prompt_id} does not exist or does not belong to user"
+ )
+
+ return prompt
+
+
+@lru_cache()
+def get_default_prompt() -> Prompt:
+ with Session(get_sqlalchemy_engine()) as db_session:
+ stmt = select(Prompt).where(Prompt.id == 0)
+
+ result = db_session.execute(stmt)
+ prompt = result.scalar_one_or_none()
+
+ if prompt is None:
+ raise RuntimeError("Default Prompt not found")
+
+ return prompt
+
+
+def get_persona_by_id(
+ persona_id: int,
+ # if user is `None` assume the user is an admin or auth is disabled
+ user: User | None,
+ db_session: Session,
+ include_deleted: bool = False,
+) -> Persona:
+ stmt = select(Persona).where(Persona.id == persona_id)
+
+ # if user is an admin, they should have access to all Personas
+ if user is not None and user.role != UserRole.ADMIN:
+ stmt = stmt.where(or_(Persona.user_id == user.id, Persona.user_id.is_(None)))
+
+ if not include_deleted:
+ stmt = stmt.where(Persona.deleted.is_(False))
+
+ result = db_session.execute(stmt)
+ persona = result.scalar_one_or_none()
+
+ if persona is None:
+ raise ValueError(
+ f"Persona with ID {persona_id} does not exist or does not belong to user"
+ )
+
+ return persona
+
+
+def get_personas_by_ids(
+ persona_ids: list[int], db_session: Session
+) -> Sequence[Persona]:
+ """Unsafe, can fetch personas from all users"""
+ if not persona_ids:
+ return []
+ personas = db_session.scalars(
+ select(Persona).where(Persona.id.in_(persona_ids))
+ ).all()
+
+ return personas
+
+
+def get_prompt_by_name(
+ prompt_name: str, user: User | None, db_session: Session
+) -> Prompt | None:
+ stmt = select(Prompt).where(Prompt.name == prompt_name)
+
+ # if user is not specified OR they are an admin, they should
+ # have access to all prompts, so this where clause is not needed
+ if user and user.role != UserRole.ADMIN:
+ stmt = stmt.where(Prompt.user_id == user.id)
+
+ result = db_session.execute(stmt).scalar_one_or_none()
+ return result
+
+
+def get_persona_by_name(
+ persona_name: str, user: User | None, db_session: Session
+) -> Persona | None:
+ """Admins can see all, regular users can only fetch their own.
+ If user is None, assume the user is an admin or auth is disabled."""
+ stmt = select(Persona).where(Persona.name == persona_name)
+ if user and user.role != UserRole.ADMIN:
+ stmt = stmt.where(Persona.user_id == user.id)
+ result = db_session.execute(stmt).scalar_one_or_none()
+ return result
+
+
+def delete_persona_by_name(
+ persona_name: str, db_session: Session, is_default: bool = True
+) -> None:
+ stmt = delete(Persona).where(
+ Persona.name == persona_name, Persona.default_persona == is_default
+ )
+
+ db_session.execute(stmt)
+
+ db_session.commit()
diff --git a/backend/danswer/db/slack_bot_config.py b/backend/danswer/db/slack_bot_config.py
index 973d7624437..43418f6217e 100644
--- a/backend/danswer/db/slack_bot_config.py
+++ b/backend/danswer/db/slack_bot_config.py
@@ -4,7 +4,6 @@
from sqlalchemy.orm import Session
from danswer.configs.chat_configs import MAX_CHUNKS_FED_TO_CHAT
-from danswer.db.chat import upsert_persona
from danswer.db.constants import SLACK_BOT_PERSONA_PREFIX
from danswer.db.document_set import get_document_sets_by_ids
from danswer.db.models import ChannelConfig
@@ -12,6 +11,10 @@
from danswer.db.models import Persona__DocumentSet
from danswer.db.models import SlackBotConfig
from danswer.db.models import SlackBotResponseType
+from danswer.db.models import User
+from danswer.db.persona import get_default_prompt
+from danswer.db.persona import mark_persona_as_deleted
+from danswer.db.persona import upsert_persona
from danswer.search.enums import RecencyBiasSetting
@@ -48,6 +51,7 @@ def create_slack_bot_persona(
# create/update persona associated with the slack bot
persona_name = _build_persona_name(channel_names)
+ default_prompt = get_default_prompt()
persona = upsert_persona(
user=None, # Slack Bot Personas are not attached to users
persona_id=existing_persona_id,
@@ -57,7 +61,7 @@ def create_slack_bot_persona(
llm_relevance_filter=True,
llm_filter_extraction=True,
recency_bias=RecencyBiasSetting.AUTO,
- prompts=None,
+ prompts=[default_prompt],
document_sets=document_sets,
llm_model_provider_override=None,
llm_model_version_override=None,
@@ -133,6 +137,7 @@ def update_slack_bot_config(
def remove_slack_bot_config(
slack_bot_config_id: int,
+ user: User | None,
db_session: Session,
) -> None:
slack_bot_config = db_session.scalar(
@@ -156,7 +161,9 @@ def remove_slack_bot_config(
_cleanup_relationships(
db_session=db_session, persona_id=existing_persona_id
)
- db_session.delete(existing_persona)
+ mark_persona_as_deleted(
+ persona_id=existing_persona_id, user=user, db_session=db_session
+ )
db_session.delete(slack_bot_config)
db_session.commit()
diff --git a/backend/danswer/llm/answering/prompts/citations_prompt.py b/backend/danswer/llm/answering/prompts/citations_prompt.py
index 81626e272c8..0a6e2c75e53 100644
--- a/backend/danswer/llm/answering/prompts/citations_prompt.py
+++ b/backend/danswer/llm/answering/prompts/citations_prompt.py
@@ -4,8 +4,8 @@
from danswer.chat.models import LlmDoc
from danswer.configs.chat_configs import MULTILINGUAL_QUERY_EXPANSION
from danswer.configs.model_configs import GEN_AI_SINGLE_USER_MESSAGE_EXPECTED_MAX_TOKENS
-from danswer.db.chat import get_default_prompt
from danswer.db.models import Persona
+from danswer.db.persona import get_default_prompt
from danswer.file_store.utils import InMemoryChatFile
from danswer.llm.answering.models import PromptConfig
from danswer.llm.factory import get_llm_for_persona
diff --git a/backend/danswer/llm/answering/stream_processing/quotes_processing.py b/backend/danswer/llm/answering/stream_processing/quotes_processing.py
index 61d379a5389..10d15b7195c 100644
--- a/backend/danswer/llm/answering/stream_processing/quotes_processing.py
+++ b/backend/danswer/llm/answering/stream_processing/quotes_processing.py
@@ -247,6 +247,17 @@ def process_model_tokens(
if found_answer_start and not found_answer_end:
if is_json_prompt and _stream_json_answer_end(model_previous, token):
found_answer_end = True
+
+ # return the remaining part of the answer e.g. token might be 'd.", ' and we should yield 'd.'
+ if token:
+ try:
+ answer_token_section = token.index('"')
+ yield DanswerAnswerPiece(
+ answer_piece=hold_quote + token[:answer_token_section]
+ )
+ except ValueError:
+ logger.error("Quotation mark not found in token")
+ yield DanswerAnswerPiece(answer_piece=hold_quote + token)
yield DanswerAnswerPiece(answer_piece=None)
continue
elif not is_json_prompt:
diff --git a/backend/danswer/llm/chat_llm.py b/backend/danswer/llm/chat_llm.py
index 4d7409eff48..d450efa9107 100644
--- a/backend/danswer/llm/chat_llm.py
+++ b/backend/danswer/llm/chat_llm.py
@@ -283,7 +283,7 @@ def _completion(
# actual input
messages=prompt,
tools=tools,
- tool_choice=tool_choice,
+ tool_choice=tool_choice if tools else None,
# streaming choice
stream=stream,
# model params
diff --git a/backend/danswer/llm/interfaces.py b/backend/danswer/llm/interfaces.py
index 917da56f2fb..1f99383fae8 100644
--- a/backend/danswer/llm/interfaces.py
+++ b/backend/danswer/llm/interfaces.py
@@ -21,12 +21,6 @@ class LLMConfig(BaseModel):
api_key: str | None
-class LLMConfig(BaseModel):
- model_provider: str
- model_name: str
- temperature: float
-
-
class LLM(abc.ABC):
"""Mimics the LangChain LLM / BaseChatModel interfaces to make it easy
to use these implementations to connect to a variety of LLM providers."""
diff --git a/backend/danswer/main.py b/backend/danswer/main.py
index a40531a527e..92caa608f87 100644
--- a/backend/danswer/main.py
+++ b/backend/danswer/main.py
@@ -27,13 +27,13 @@
from danswer.configs.app_configs import AUTH_TYPE
from danswer.configs.app_configs import DISABLE_GENERATIVE_AI
from danswer.configs.app_configs import DISABLE_INDEX_UPDATE_ON_SWAP
+from danswer.configs.app_configs import LOG_ENDPOINT_LATENCY
from danswer.configs.app_configs import OAUTH_CLIENT_ID
from danswer.configs.app_configs import OAUTH_CLIENT_SECRET
from danswer.configs.app_configs import USER_AUTH_SECRET
from danswer.configs.app_configs import WEB_DOMAIN
from danswer.configs.chat_configs import MULTILINGUAL_QUERY_EXPANSION
from danswer.configs.constants import AuthType
-from danswer.db.chat import delete_old_default_personas
from danswer.db.connector import create_initial_default_connector
from danswer.db.connector_credential_pair import associate_default_cc_pair
from danswer.db.connector_credential_pair import get_connector_credential_pairs
@@ -42,8 +42,10 @@
from danswer.db.embedding_model import get_current_db_embedding_model
from danswer.db.embedding_model import get_secondary_db_embedding_model
from danswer.db.engine import get_sqlalchemy_engine
+from danswer.db.engine import warm_up_connections
from danswer.db.index_attempt import cancel_indexing_attempts_past_model
from danswer.db.index_attempt import expire_index_attempts
+from danswer.db.persona import delete_old_default_personas
from danswer.db.swap_index import check_index_swap
from danswer.document_index.factory import get_default_document_index
from danswer.llm.llm_initialization import load_llm_providers
@@ -70,6 +72,7 @@
from danswer.server.manage.secondary_index import router as secondary_index_router
from danswer.server.manage.slack_bot import router as slack_bot_management_router
from danswer.server.manage.users import router as user_router
+from danswer.server.middleware.latency_logging import add_latency_logging_middleware
from danswer.server.query_and_chat.chat_backend import router as chat_router
from danswer.server.query_and_chat.query_backend import (
admin_router as admin_query_router,
@@ -165,6 +168,9 @@ async def lifespan(app: FastAPI) -> AsyncGenerator:
f"Using multilingual flow with languages: {MULTILINGUAL_QUERY_EXPANSION}"
)
+ # fill up Postgres connection pools
+ await warm_up_connections()
+
with Session(engine) as db_session:
check_index_swap(db_session=db_session)
db_embedding_model = get_current_db_embedding_model(db_session)
@@ -356,6 +362,8 @@ def get_application() -> FastAPI:
allow_methods=["*"],
allow_headers=["*"],
)
+ if LOG_ENDPOINT_LATENCY:
+ add_latency_logging_middleware(application, logger)
# Ensure all routes have auth enabled or are explicitly marked as public
check_router_auth(application)
diff --git a/backend/danswer/one_shot_answer/answer_question.py b/backend/danswer/one_shot_answer/answer_question.py
index e5f45ada12d..2555d2d708e 100644
--- a/backend/danswer/one_shot_answer/answer_question.py
+++ b/backend/danswer/one_shot_answer/answer_question.py
@@ -19,11 +19,11 @@
from danswer.db.chat import create_db_search_doc
from danswer.db.chat import create_new_chat_message
from danswer.db.chat import get_or_create_root_message
-from danswer.db.chat import get_prompt_by_id
from danswer.db.chat import translate_db_message_to_chat_message_detail
from danswer.db.chat import translate_db_search_doc_to_server_search_doc
from danswer.db.engine import get_session_context_manager
from danswer.db.models import User
+from danswer.db.persona import get_prompt_by_id
from danswer.llm.answering.answer import Answer
from danswer.llm.answering.models import AnswerStyleConfig
from danswer.llm.answering.models import CitationConfig
@@ -39,6 +39,8 @@
from danswer.search.models import RerankMetricsContainer
from danswer.search.models import RetrievalMetricsContainer
from danswer.search.utils import chunks_or_sections_to_search_docs
+from danswer.search.utils import dedupe_documents
+from danswer.search.utils import drop_llm_indices
from danswer.secondary_llm_flows.answer_validation import get_answer_validity
from danswer.secondary_llm_flows.query_expansion import thread_based_query_rephrase
from danswer.server.query_and_chat.models import ChatMessageDetail
@@ -130,8 +132,10 @@ def stream_answer_objects(
prompt = None
if query_req.prompt_id is not None:
+ # NOTE: let the user access any prompt as long as the Persona is shared
+ # with them
prompt = get_prompt_by_id(
- prompt_id=query_req.prompt_id, user=user, db_session=db_session
+ prompt_id=query_req.prompt_id, user=None, db_session=db_session
)
if prompt is None:
if not chat_session.persona.prompts:
@@ -194,6 +198,7 @@ def stream_answer_objects(
skip_explicit_tool_calling=True,
)
# won't be any ImageGenerationDisplay responses since that tool is never passed in
+ dropped_inds: list[int] = []
for packet in cast(AnswerObjectIterator, answer.processed_streamed_output):
# for one-shot flow, don't currently do anything with these
if isinstance(packet, ToolResponse):
@@ -204,11 +209,14 @@ def stream_answer_objects(
search_response_summary.top_sections
)
+ # Deduping happens at the last step to avoid harming quality by dropping content early on
+ deduped_docs = top_docs
+ if query_req.retrieval_options.dedupe_docs:
+ deduped_docs, dropped_inds = dedupe_documents(top_docs)
+
reference_db_search_docs = [
- create_db_search_doc(
- server_search_doc=top_doc, db_session=db_session
- )
- for top_doc in top_docs
+ create_db_search_doc(server_search_doc=doc, db_session=db_session)
+ for doc in deduped_docs
]
response_docs = [
@@ -227,6 +235,15 @@ def stream_answer_objects(
)
yield initial_response
elif packet.id == SECTION_RELEVANCE_LIST_ID:
+ chunk_indices = packet.response
+
+ if reference_db_search_docs is not None and dropped_inds:
+ chunk_indices = drop_llm_indices(
+ llm_indices=chunk_indices,
+ search_docs=reference_db_search_docs,
+ dropped_indices=dropped_inds,
+ )
+
yield LLMRelevanceFilterResponse(relevant_chunk_indices=packet.response)
else:
yield packet
diff --git a/backend/danswer/prompts/miscellaneous_prompts.py b/backend/danswer/prompts/miscellaneous_prompts.py
index 340908f11c6..876f0f3591a 100644
--- a/backend/danswer/prompts/miscellaneous_prompts.py
+++ b/backend/danswer/prompts/miscellaneous_prompts.py
@@ -11,6 +11,20 @@
{query}
""".strip()
+SLACK_LANGUAGE_REPHRASE_PROMPT = """
+As an AI assistant employed by an organization, \
+your role is to transform user messages into concise \
+inquiries suitable for a Large Language Model (LLM) that \
+retrieves pertinent materials within a Retrieval-Augmented \
+Generation (RAG) framework. Ensure to reply in the identical \
+language as the original request. When faced with multiple \
+questions within a single query, distill them into a singular, \
+unified question, disregarding any direct mentions.
+
+Query:
+{query}
+""".strip()
+
# Use the following for easy viewing of prompts
if __name__ == "__main__":
diff --git a/backend/danswer/prompts/prompt_utils.py b/backend/danswer/prompts/prompt_utils.py
index 1e9316d2b71..9dc939eb170 100644
--- a/backend/danswer/prompts/prompt_utils.py
+++ b/backend/danswer/prompts/prompt_utils.py
@@ -39,7 +39,7 @@ def add_time_to_system_prompt(system_prompt: str) -> str:
if DANSWER_DATETIME_REPLACEMENT in system_prompt:
return system_prompt.replace(
DANSWER_DATETIME_REPLACEMENT,
- get_current_llm_day_time(full_sentence=False, include_day_of_week=False),
+ get_current_llm_day_time(full_sentence=False, include_day_of_week=True),
)
if system_prompt:
diff --git a/backend/danswer/search/models.py b/backend/danswer/search/models.py
index 16a64d82049..f0069d5c750 100644
--- a/backend/danswer/search/models.py
+++ b/backend/danswer/search/models.py
@@ -116,6 +116,9 @@ class RetrievalDetails(ChunkContext):
offset: int | None = None
limit: int | None = None
+ # If this is set, only the highest matching chunk (or merged chunks) is returned
+ dedupe_docs: bool = False
+
class InferenceChunk(BaseChunk):
document_id: str
diff --git a/backend/danswer/search/search_nlp_models.py b/backend/danswer/search/search_nlp_models.py
index 5243a29c067..761d9aa791f 100644
--- a/backend/danswer/search/search_nlp_models.py
+++ b/backend/danswer/search/search_nlp_models.py
@@ -24,6 +24,7 @@
os.environ["TOKENIZERS_PARALLELISM"] = "false"
os.environ["HF_HUB_DISABLE_TELEMETRY"] = "1"
+os.environ["TRANSFORMERS_NO_ADVISORY_WARNINGS"] = "1"
logger = setup_logger()
diff --git a/backend/danswer/search/utils.py b/backend/danswer/search/utils.py
index fbcb205e3cf..5b5a6464aad 100644
--- a/backend/danswer/search/utils.py
+++ b/backend/danswer/search/utils.py
@@ -1,10 +1,39 @@
from collections.abc import Sequence
+from typing import TypeVar
+from danswer.db.models import SearchDoc as DBSearchDoc
from danswer.search.models import InferenceChunk
from danswer.search.models import InferenceSection
from danswer.search.models import SearchDoc
+T = TypeVar("T", InferenceSection, InferenceChunk, SearchDoc)
+
+
+def dedupe_documents(items: list[T]) -> tuple[list[T], list[int]]:
+ seen_ids = set()
+ deduped_items = []
+ dropped_indices = []
+ for index, item in enumerate(items):
+ if item.document_id not in seen_ids:
+ seen_ids.add(item.document_id)
+ deduped_items.append(item)
+ else:
+ dropped_indices.append(index)
+ return deduped_items, dropped_indices
+
+
+def drop_llm_indices(
+ llm_indices: list[int], search_docs: list[DBSearchDoc], dropped_indices: list[int]
+) -> list[int]:
+ llm_bools = [True if i in llm_indices else False for i in range(len(search_docs))]
+ if dropped_indices:
+ llm_bools = [
+ val for ind, val in enumerate(llm_bools) if ind not in dropped_indices
+ ]
+ return [i for i, val in enumerate(llm_bools) if val]
+
+
def chunks_or_sections_to_search_docs(
chunks: Sequence[InferenceChunk | InferenceSection] | None,
) -> list[SearchDoc]:
diff --git a/backend/danswer/secondary_llm_flows/choose_search.py b/backend/danswer/secondary_llm_flows/choose_search.py
index df3597d5641..5016cf055bc 100644
--- a/backend/danswer/secondary_llm_flows/choose_search.py
+++ b/backend/danswer/secondary_llm_flows/choose_search.py
@@ -3,6 +3,7 @@
from langchain.schema import SystemMessage
from danswer.chat.chat_utils import combine_message_chain
+from danswer.configs.chat_configs import DISABLE_LLM_CHOOSE_SEARCH
from danswer.configs.model_configs import GEN_AI_HISTORY_CUTOFF
from danswer.db.models import ChatMessage
from danswer.llm.answering.models import PreviousMessage
@@ -26,8 +27,8 @@ def check_if_need_search_multi_message(
history: list[ChatMessage],
llm: LLM,
) -> bool:
- # Always start with a retrieval
- if not history:
+ # Retrieve on start or when choosing is globally disabled
+ if not history or DISABLE_LLM_CHOOSE_SEARCH:
return True
prompt_msgs: list[BaseMessage] = [SystemMessage(content=REQUIRE_SEARCH_SYSTEM_MSG)]
@@ -65,6 +66,10 @@ def _get_search_messages(
return messages
+ # Choosing is globally disabled, use search
+ if DISABLE_LLM_CHOOSE_SEARCH:
+ return True
+
history_str = combine_message_chain(
messages=history, token_limit=GEN_AI_HISTORY_CUTOFF
)
diff --git a/backend/danswer/server/auth_check.py b/backend/danswer/server/auth_check.py
index 9fbc973bdc6..53ef572daa3 100644
--- a/backend/danswer/server/auth_check.py
+++ b/backend/danswer/server/auth_check.py
@@ -17,7 +17,7 @@
("/docs/oauth2-redirect", {"GET", "HEAD"}),
("/redoc", {"GET", "HEAD"}),
# should always be callable, will just return 401 if not authenticated
- ("/manage/me", {"GET"}),
+ ("/me", {"GET"}),
# just returns 200 to validate that the server is up
("/health", {"GET"}),
# just returns auth type, needs to be accessible before the user is logged
diff --git a/backend/danswer/server/features/persona/api.py b/backend/danswer/server/features/persona/api.py
index 65303e4b782..006f7506894 100644
--- a/backend/danswer/server/features/persona/api.py
+++ b/backend/danswer/server/features/persona/api.py
@@ -1,3 +1,5 @@
+from uuid import UUID
+
from fastapi import APIRouter
from fastapi import Depends
from pydantic import BaseModel
@@ -5,15 +7,16 @@
from danswer.auth.users import current_admin_user
from danswer.auth.users import current_user
-from danswer.db.chat import get_persona_by_id
-from danswer.db.chat import get_personas
-from danswer.db.chat import mark_persona_as_deleted
-from danswer.db.chat import mark_persona_as_not_deleted
-from danswer.db.chat import update_all_personas_display_priority
-from danswer.db.chat import update_persona_visibility
from danswer.db.engine import get_session
from danswer.db.models import User
from danswer.db.persona import create_update_persona
+from danswer.db.persona import get_persona_by_id
+from danswer.db.persona import get_personas
+from danswer.db.persona import mark_persona_as_deleted
+from danswer.db.persona import mark_persona_as_not_deleted
+from danswer.db.persona import update_all_personas_display_priority
+from danswer.db.persona import update_persona_shared_users
+from danswer.db.persona import update_persona_visibility
from danswer.llm.answering.prompts.utils import build_dummy_prompt
from danswer.server.features.persona.models import CreatePersonaRequest
from danswer.server.features.persona.models import PersonaSnapshot
@@ -119,6 +122,25 @@ def update_persona(
)
+class PersonaShareRequest(BaseModel):
+ user_ids: list[UUID]
+
+
+@basic_router.patch("/{persona_id}/share")
+def share_persona(
+ persona_id: int,
+ persona_share_request: PersonaShareRequest,
+ user: User | None = Depends(current_user),
+ db_session: Session = Depends(get_session),
+) -> None:
+ update_persona_shared_users(
+ persona_id=persona_id,
+ user_ids=persona_share_request.user_ids,
+ user=user,
+ db_session=db_session,
+ )
+
+
@basic_router.delete("/{persona_id}")
def delete_persona(
persona_id: int,
diff --git a/backend/danswer/server/features/persona/models.py b/backend/danswer/server/features/persona/models.py
index 46ccba760aa..aee39e72af0 100644
--- a/backend/danswer/server/features/persona/models.py
+++ b/backend/danswer/server/features/persona/models.py
@@ -53,7 +53,7 @@ class PersonaSnapshot(BaseModel):
prompts: list[PromptSnapshot]
tools: list[ToolSnapshot]
document_sets: list[DocumentSet]
- users: list[UUID]
+ users: list[MinimalUserSnapshot]
groups: list[int]
@classmethod
@@ -92,7 +92,10 @@ def from_model(
DocumentSet.from_model(document_set_model)
for document_set_model in persona.document_sets
],
- users=[user.id for user in persona.users],
+ users=[
+ MinimalUserSnapshot(id=user.id, email=user.email)
+ for user in persona.users
+ ],
groups=[user_group.id for user_group in persona.groups],
)
diff --git a/backend/danswer/server/features/prompt/api.py b/backend/danswer/server/features/prompt/api.py
index 24c886ab915..aebcbb8434d 100644
--- a/backend/danswer/server/features/prompt/api.py
+++ b/backend/danswer/server/features/prompt/api.py
@@ -5,13 +5,13 @@
from starlette import status
from danswer.auth.users import current_user
-from danswer.db.chat import get_personas_by_ids
-from danswer.db.chat import get_prompt_by_id
-from danswer.db.chat import get_prompts
-from danswer.db.chat import mark_prompt_as_deleted
-from danswer.db.chat import upsert_prompt
from danswer.db.engine import get_session
from danswer.db.models import User
+from danswer.db.persona import get_personas_by_ids
+from danswer.db.persona import get_prompt_by_id
+from danswer.db.persona import get_prompts
+from danswer.db.persona import mark_prompt_as_deleted
+from danswer.db.persona import upsert_prompt
from danswer.server.features.prompt.models import CreatePromptRequest
from danswer.server.features.prompt.models import PromptSnapshot
from danswer.utils.logger import setup_logger
diff --git a/backend/danswer/server/manage/llm/api.py b/backend/danswer/server/manage/llm/api.py
index 7bc4efe63cc..6e21a29b12b 100644
--- a/backend/danswer/server/manage/llm/api.py
+++ b/backend/danswer/server/manage/llm/api.py
@@ -55,13 +55,13 @@ def test_llm_configuration(
functions_with_args: list[tuple[Callable, tuple]] = [(test_llm, (llm,))]
if (
- test_llm_request.default_fast_model_name
- and test_llm_request.default_fast_model_name
+ test_llm_request.fast_default_model_name
+ and test_llm_request.fast_default_model_name
!= test_llm_request.default_model_name
):
fast_llm = get_llm(
provider=test_llm_request.provider,
- model=test_llm_request.default_fast_model_name,
+ model=test_llm_request.fast_default_model_name,
api_key=test_llm_request.api_key,
api_base=test_llm_request.api_base,
api_version=test_llm_request.api_version,
diff --git a/backend/danswer/server/manage/llm/models.py b/backend/danswer/server/manage/llm/models.py
index 0e791696a76..05a596ffd54 100644
--- a/backend/danswer/server/manage/llm/models.py
+++ b/backend/danswer/server/manage/llm/models.py
@@ -18,7 +18,7 @@ class TestLLMRequest(BaseModel):
# model level
default_model_name: str
- default_fast_model_name: str | None = None
+ fast_default_model_name: str | None = None
class LLMProviderDescriptor(BaseModel):
diff --git a/backend/danswer/server/manage/models.py b/backend/danswer/server/manage/models.py
index df0f2ec477c..8797913f44f 100644
--- a/backend/danswer/server/manage/models.py
+++ b/backend/danswer/server/manage/models.py
@@ -1,4 +1,5 @@
from typing import Any
+from typing import TYPE_CHECKING
from pydantic import BaseModel
from pydantic import root_validator
@@ -14,6 +15,9 @@
from danswer.indexing.models import EmbeddingModelDetail
from danswer.server.features.persona.models import PersonaSnapshot
+if TYPE_CHECKING:
+ from danswer.db.models import User as UserModel
+
class VersionResponse(BaseModel):
backend_version: str
@@ -26,6 +30,10 @@ class AuthTypeResponse(BaseModel):
requires_verification: bool
+class UserPreferences(BaseModel):
+ chosen_assistants: list[int] | None
+
+
class UserInfo(BaseModel):
id: str
email: str
@@ -33,6 +41,19 @@ class UserInfo(BaseModel):
is_superuser: bool
is_verified: bool
role: UserRole
+ preferences: UserPreferences
+
+ @classmethod
+ def from_model(cls, user: "UserModel") -> "UserInfo":
+ return cls(
+ id=str(user.id),
+ email=user.email,
+ is_active=user.is_active,
+ is_superuser=user.is_superuser,
+ is_verified=user.is_verified,
+ role=user.role,
+ preferences=(UserPreferences(chosen_assistants=user.chosen_assistants)),
+ )
class UserByEmail(BaseModel):
diff --git a/backend/danswer/server/manage/slack_bot.py b/backend/danswer/server/manage/slack_bot.py
index 40e8663b054..aade420adeb 100644
--- a/backend/danswer/server/manage/slack_bot.py
+++ b/backend/danswer/server/manage/slack_bot.py
@@ -7,11 +7,11 @@
from danswer.danswerbot.slack.config import validate_channel_names
from danswer.danswerbot.slack.tokens import fetch_tokens
from danswer.danswerbot.slack.tokens import save_tokens
-from danswer.db.chat import get_persona_by_id
from danswer.db.constants import SLACK_BOT_PERSONA_PREFIX
from danswer.db.engine import get_session
from danswer.db.models import ChannelConfig
from danswer.db.models import User
+from danswer.db.persona import get_persona_by_id
from danswer.db.slack_bot_config import create_slack_bot_persona
from danswer.db.slack_bot_config import fetch_slack_bot_config
from danswer.db.slack_bot_config import fetch_slack_bot_configs
@@ -172,10 +172,10 @@ def patch_slack_bot_config(
def delete_slack_bot_config(
slack_bot_config_id: int,
db_session: Session = Depends(get_session),
- _: User | None = Depends(current_admin_user),
+ user: User | None = Depends(current_admin_user),
) -> None:
remove_slack_bot_config(
- slack_bot_config_id=slack_bot_config_id, db_session=db_session
+ slack_bot_config_id=slack_bot_config_id, user=user, db_session=db_session
)
diff --git a/backend/danswer/server/manage/users.py b/backend/danswer/server/manage/users.py
index 7258a8564f3..a643f4e752f 100644
--- a/backend/danswer/server/manage/users.py
+++ b/backend/danswer/server/manage/users.py
@@ -2,26 +2,33 @@
from fastapi import Depends
from fastapi import HTTPException
from fastapi import status
+from pydantic import BaseModel
+from sqlalchemy import update
from sqlalchemy.orm import Session
+from danswer.auth.noauth_user import fetch_no_auth_user
+from danswer.auth.noauth_user import set_no_auth_user_preferences
from danswer.auth.schemas import UserRead
from danswer.auth.schemas import UserRole
from danswer.auth.users import current_admin_user
from danswer.auth.users import current_user
-from danswer.auth.users import get_display_email
from danswer.auth.users import optional_user
+from danswer.configs.app_configs import AUTH_TYPE
+from danswer.configs.constants import AuthType
from danswer.db.engine import get_session
from danswer.db.models import User
from danswer.db.users import get_user_by_email
from danswer.db.users import list_users
+from danswer.dynamic_configs.factory import get_dynamic_config_store
from danswer.server.manage.models import UserByEmail
from danswer.server.manage.models import UserInfo
from danswer.server.manage.models import UserRoleResponse
+from danswer.server.models import MinimalUserSnapshot
-router = APIRouter(prefix="/manage")
+router = APIRouter()
-@router.patch("/promote-user-to-admin")
+@router.patch("/manage/promote-user-to-admin")
def promote_admin(
user_email: UserByEmail,
_: User = Depends(current_admin_user),
@@ -38,7 +45,7 @@ def promote_admin(
db_session.commit()
-@router.patch("/demote-admin-to-basic")
+@router.patch("/manage/demote-admin-to-basic")
async def demote_admin(
user_email: UserByEmail,
user: User = Depends(current_admin_user),
@@ -60,7 +67,7 @@ async def demote_admin(
db_session.commit()
-@router.get("/users")
+@router.get("/manage/users")
def list_all_users(
_: User | None = Depends(current_admin_user),
db_session: Session = Depends(get_session),
@@ -69,7 +76,19 @@ def list_all_users(
return [UserRead.from_orm(user) for user in users]
-@router.get("/get-user-role", response_model=UserRoleResponse)
+"""Endpoints for all"""
+
+
+@router.get("/users")
+def list_all_users_basic_info(
+ _: User | None = Depends(current_user),
+ db_session: Session = Depends(get_session),
+) -> list[MinimalUserSnapshot]:
+ users = list_users(db_session)
+ return [MinimalUserSnapshot(id=user.id, email=user.email) for user in users]
+
+
+@router.get("/get-user-role")
async def get_user_role(user: User = Depends(current_user)) -> UserRoleResponse:
if user is None:
raise ValueError("Invalid or missing user.")
@@ -77,20 +96,53 @@ async def get_user_role(user: User = Depends(current_user)) -> UserRoleResponse:
@router.get("/me")
-def verify_user_logged_in(user: User | None = Depends(optional_user)) -> UserInfo:
+def verify_user_logged_in(
+ user: User | None = Depends(optional_user),
+) -> UserInfo:
# NOTE: this does not use `current_user` / `current_admin_user` because we don't want
# to enforce user verification here - the frontend always wants to get the info about
# the current user regardless of if they are currently verified
if user is None:
+ # if auth type is disabled, return a dummy user with preferences from
+ # the key-value store
+ if AUTH_TYPE == AuthType.DISABLED:
+ store = get_dynamic_config_store()
+ return fetch_no_auth_user(store)
+
raise HTTPException(
status_code=status.HTTP_403_FORBIDDEN, detail="User Not Authenticated"
)
- return UserInfo(
- id=str(user.id),
- email=get_display_email(user.email, space_less=True),
- is_active=user.is_active,
- is_superuser=user.is_superuser,
- is_verified=user.is_verified,
- role=user.role,
+ return UserInfo.from_model(user)
+
+
+"""APIs to adjust user preferences"""
+
+
+class ChosenAssistantsRequest(BaseModel):
+ chosen_assistants: list[int]
+
+
+@router.patch("/user/assistant-list")
+def update_user_assistant_list(
+ request: ChosenAssistantsRequest,
+ user: User | None = Depends(current_user),
+ db_session: Session = Depends(get_session),
+) -> None:
+ if user is None:
+ if AUTH_TYPE == AuthType.DISABLED:
+ store = get_dynamic_config_store()
+
+ no_auth_user = fetch_no_auth_user(store)
+ no_auth_user.preferences.chosen_assistants = request.chosen_assistants
+ set_no_auth_user_preferences(store, no_auth_user.preferences)
+ return
+ else:
+ raise RuntimeError("This should never happen")
+
+ db_session.execute(
+ update(User)
+ .where(User.id == user.id) # type: ignore
+ .values(chosen_assistants=request.chosen_assistants)
)
+ db_session.commit()
diff --git a/backend/danswer/server/middleware/latency_logging.py b/backend/danswer/server/middleware/latency_logging.py
new file mode 100644
index 00000000000..f2bc3127af4
--- /dev/null
+++ b/backend/danswer/server/middleware/latency_logging.py
@@ -0,0 +1,23 @@
+import logging
+import time
+from collections.abc import Awaitable
+from collections.abc import Callable
+
+from fastapi import FastAPI
+from fastapi import Request
+from fastapi import Response
+
+
+def add_latency_logging_middleware(app: FastAPI, logger: logging.LoggerAdapter) -> None:
+ @app.middleware("http")
+ async def log_latency(
+ request: Request, call_next: Callable[[Request], Awaitable[Response]]
+ ) -> Response:
+ start_time = time.monotonic()
+ response = await call_next(request)
+ process_time = time.monotonic() - start_time
+ logger.info(
+ f"Path: {request.url.path} - Method: {request.method} - "
+ f"Status Code: {response.status_code} - Time: {process_time:.4f} secs"
+ )
+ return response
diff --git a/backend/danswer/server/query_and_chat/chat_backend.py b/backend/danswer/server/query_and_chat/chat_backend.py
index 97dcc62d00e..ab72699c2bc 100644
--- a/backend/danswer/server/query_and_chat/chat_backend.py
+++ b/backend/danswer/server/query_and_chat/chat_backend.py
@@ -24,7 +24,6 @@
from danswer.db.chat import get_chat_session_by_id
from danswer.db.chat import get_chat_sessions_by_user
from danswer.db.chat import get_or_create_root_message
-from danswer.db.chat import get_persona_by_id
from danswer.db.chat import set_as_latest_chat_message
from danswer.db.chat import translate_db_message_to_chat_message_detail
from danswer.db.chat import update_chat_session
@@ -32,6 +31,7 @@
from danswer.db.feedback import create_chat_message_feedback
from danswer.db.feedback import create_doc_retrieval_feedback
from danswer.db.models import User
+from danswer.db.persona import get_persona_by_id
from danswer.document_index.document_index_utils import get_both_index_names
from danswer.document_index.factory import get_default_document_index
from danswer.file_processing.extract_file_text import extract_file_text
diff --git a/backend/requirements/default.txt b/backend/requirements/default.txt
index 710fa0d3ba1..bcf8478f85f 100644
--- a/backend/requirements/default.txt
+++ b/backend/requirements/default.txt
@@ -35,7 +35,7 @@ llama-index==0.9.45
Mako==1.2.4
msal==1.26.0
nltk==3.8.1
-Office365-REST-Python-Client==2.5.4
+Office365-REST-Python-Client==2.5.9
oauthlib==3.2.2
openai==1.14.3
openpyxl==3.1.2
@@ -71,3 +71,4 @@ uvicorn==0.21.1
zulip==0.8.2
hubspot-api-client==8.1.0
zenpy==2.0.41
+dropbox==11.36.2
diff --git a/deployment/data/nginx/app.conf.template b/deployment/data/nginx/app.conf.template
index e02d8ff2f84..b698c744bf5 100644
--- a/deployment/data/nginx/app.conf.template
+++ b/deployment/data/nginx/app.conf.template
@@ -1,3 +1,9 @@
+# Log format to include request latency
+log_format custom_main '$remote_addr - $remote_user [$time_local] "$request" '
+ '$status $body_bytes_sent "$http_referer" '
+ '"$http_user_agent" "$http_x_forwarded_for" '
+ 'rt=$request_time';
+
upstream api_server {
# fail_timeout=0 means we always retry an upstream even if it failed
# to return a good HTTP response
@@ -20,6 +26,8 @@ server {
client_max_body_size 5G; # Maximum upload size
+ access_log /var/log/nginx/access.log custom_main;
+
# Match both /api/* and /openapi.json in a single rule
location ~ ^/(api|openapi.json)(/.*)?$ {
# Rewrite /api prefixed matched paths
diff --git a/deployment/data/nginx/app.conf.template.dev b/deployment/data/nginx/app.conf.template.dev
index a0e7237e7e8..a7a0efa192b 100644
--- a/deployment/data/nginx/app.conf.template.dev
+++ b/deployment/data/nginx/app.conf.template.dev
@@ -1,3 +1,9 @@
+# Override log format to include request latency
+log_format custom_main '$remote_addr - $remote_user [$time_local] "$request" '
+ '$status $body_bytes_sent "$http_referer" '
+ '"$http_user_agent" "$http_x_forwarded_for" '
+ 'rt=$request_time';
+
upstream api_server {
# fail_timeout=0 means we always retry an upstream even if it failed
# to return a good HTTP response
@@ -20,6 +26,8 @@ server {
client_max_body_size 5G; # Maximum upload size
+ access_log /var/log/nginx/access.log custom_main;
+
# Match both /api/* and /openapi.json in a single rule
location ~ ^/(api|openapi.json)(/.*)?$ {
# Rewrite /api prefixed matched paths
@@ -58,3 +66,4 @@ server {
proxy_pass http://web_server;
}
}
+
diff --git a/deployment/data/nginx/app.conf.template.no-letsencrypt b/deployment/data/nginx/app.conf.template.no-letsencrypt
index abf4371fc8e..4d5096374a4 100644
--- a/deployment/data/nginx/app.conf.template.no-letsencrypt
+++ b/deployment/data/nginx/app.conf.template.no-letsencrypt
@@ -1,3 +1,9 @@
+# Log format to include request latency
+log_format custom_main '$remote_addr - $remote_user [$time_local] "$request" '
+ '$status $body_bytes_sent "$http_referer" '
+ '"$http_user_agent" "$http_x_forwarded_for" '
+ 'rt=$request_time';
+
upstream api_server {
# fail_timeout=0 means we always retry an upstream even if it failed
# to return a good HTTP response
@@ -20,6 +26,8 @@ server {
client_max_body_size 5G; # Maximum upload size
+ access_log /var/log/nginx/access.log custom_main;
+
# Match both /api/* and /openapi.json in a single rule
location ~ ^/(api|openapi.json)(/.*)?$ {
# Rewrite /api prefixed matched paths
diff --git a/deployment/docker_compose/README.md b/deployment/docker_compose/README.md
index 96285681e94..a12f22bea7d 100644
--- a/deployment/docker_compose/README.md
+++ b/deployment/docker_compose/README.md
@@ -2,10 +2,10 @@
# Deploying Danswer using Docker Compose
-For general information, please read the instructions in this [README](https://github.com/danswer-ai/danswer/blob/main/deployment/docker_compose/README.md).
+For general information, please read the instructions in this [README](https://github.com/danswer-ai/danswer/blob/main/deployment/README.md).
## Deploy in a system without GPU support
-This part is elaborated precisely in in this [README](https://github.com/danswer-ai/danswer/blob/main/deployment/docker_compose/README.md) in section *Docker Compose*. If you have any questions, please feel free to open an issue or get in touch in slack for support.
+This part is elaborated precisely in in this [README](https://github.com/danswer-ai/danswer/blob/main/deployment/README.md) in section *Docker Compose*. If you have any questions, please feel free to open an issue or get in touch in slack for support.
## Deploy in a system with GPU support
Running Model servers with GPU support while indexing and querying can result in significant improvements in performance. This is highly recommended if you have access to resources. Currently, Danswer offloads embedding model and tokenizers to the GPU VRAM and the size needed depends on chosen embedding model. Default embedding models `intfloat/e5-base-v2` takes up about 1GB of VRAM and since we need this for inference and embedding pipeline, you would need roughly 2GB of VRAM.
diff --git a/deployment/docker_compose/docker-compose.dev.yml b/deployment/docker_compose/docker-compose.dev.yml
index 7dcc3bc347d..c7175faa1f5 100644
--- a/deployment/docker_compose/docker-compose.dev.yml
+++ b/deployment/docker_compose/docker-compose.dev.yml
@@ -81,6 +81,7 @@ services:
# If set to `true` will enable additional logs about Vespa query performance
# (time spent on finding the right docs + time spent fetching summaries from disk)
- LOG_VESPA_TIMING_INFORMATION=${LOG_VESPA_TIMING_INFORMATION:-}
+ - LOG_ENDPOINT_LATENCY=${LOG_ENDPOINT_LATENCY:-}
extra_hosts:
- "host.docker.internal:host-gateway"
logging:
@@ -195,6 +196,7 @@ services:
- NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA=${NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA:-false}
- NEXT_PUBLIC_POSITIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_POSITIVE_PREDEFINED_FEEDBACK_OPTIONS:-}
- NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS:-}
+ - NEXT_PUBLIC_DISABLE_LOGOUT=${NEXT_PUBLIC_DISABLE_LOGOUT:-}
depends_on:
- api_server
restart: always
diff --git a/deployment/docker_compose/docker-compose.gpu-dev.yml b/deployment/docker_compose/docker-compose.gpu-dev.yml
index f6c2c7fdd30..89dd673f91b 100644
--- a/deployment/docker_compose/docker-compose.gpu-dev.yml
+++ b/deployment/docker_compose/docker-compose.gpu-dev.yml
@@ -197,6 +197,7 @@ services:
- NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA=${NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA:-false}
- NEXT_PUBLIC_POSITIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_POSITIVE_PREDEFINED_FEEDBACK_OPTIONS:-}
- NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS:-}
+ - NEXT_PUBLIC_DISABLE_LOGOUT=${NEXT_PUBLIC_DISABLE_LOGOUT:-}
depends_on:
- api_server
restart: always
@@ -214,6 +215,7 @@ services:
reservations:
devices:
- driver: nvidia
+ count: all
capabilities: [gpu]
build:
context: ../../backend
@@ -252,6 +254,7 @@ services:
reservations:
devices:
- driver: nvidia
+ count: all
capabilities: [gpu]
command: >
/bin/sh -c "if [ \"${DISABLE_MODEL_SERVER:-false}\" = \"True\" ]; then
diff --git a/deployment/docker_compose/docker-compose.prod-no-letsencrypt.yml b/deployment/docker_compose/docker-compose.prod-no-letsencrypt.yml
index 226df69d8e4..c2ef8a13274 100644
--- a/deployment/docker_compose/docker-compose.prod-no-letsencrypt.yml
+++ b/deployment/docker_compose/docker-compose.prod-no-letsencrypt.yml
@@ -67,6 +67,7 @@ services:
- NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA=${NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA:-false}
- NEXT_PUBLIC_POSITIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_POSITIVE_PREDEFINED_FEEDBACK_OPTIONS:-}
- NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS:-}
+ - NEXT_PUBLIC_DISABLE_LOGOUT=${NEXT_PUBLIC_DISABLE_LOGOUT:-}
depends_on:
- api_server
restart: always
diff --git a/deployment/docker_compose/docker-compose.prod.yml b/deployment/docker_compose/docker-compose.prod.yml
index b93ce7c3843..26ef6101ce7 100644
--- a/deployment/docker_compose/docker-compose.prod.yml
+++ b/deployment/docker_compose/docker-compose.prod.yml
@@ -67,6 +67,7 @@ services:
- NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA=${NEXT_PUBLIC_NEW_CHAT_DIRECTS_TO_SAME_PERSONA:-false}
- NEXT_PUBLIC_POSITIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_POSITIVE_PREDEFINED_FEEDBACK_OPTIONS:-}
- NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS:-}
+ - NEXT_PUBLIC_DISABLE_LOGOUT=${NEXT_PUBLIC_DISABLE_LOGOUT:-}
depends_on:
- api_server
restart: always
diff --git a/deployment/docker_compose/init-letsencrypt.sh b/deployment/docker_compose/init-letsencrypt.sh
index 5eb3c73b9a1..9eec409fada 100755
--- a/deployment/docker_compose/init-letsencrypt.sh
+++ b/deployment/docker_compose/init-letsencrypt.sh
@@ -21,7 +21,13 @@ docker_compose_cmd() {
# Assign appropriate Docker Compose command
COMPOSE_CMD=$(docker_compose_cmd)
-domains=("$DOMAIN" "www.$DOMAIN")
+# Only add www to domain list if domain wasn't explicitly set as a subdomain
+if [[ ! $DOMAIN == www.* ]]; then
+ domains=("$DOMAIN" "www.$DOMAIN")
+else
+ domains=("$DOMAIN")
+fi
+
rsa_key_size=4096
data_path="../data/certbot"
email="$EMAIL" # Adding a valid address is strongly recommended
diff --git a/deployment/helm/Chart.lock b/deployment/helm/Chart.lock
index 7486bf317f2..918b44f6ebf 100644
--- a/deployment/helm/Chart.lock
+++ b/deployment/helm/Chart.lock
@@ -1,6 +1,12 @@
dependencies:
- name: postgresql
repository: https://charts.bitnami.com/bitnami
- version: 14.1.0
-digest: sha256:526d286ca7143959104d8a7f9b196706efdbd89dcc37943a1b54016f224d4b4d
-generated: "2024-02-16T12:21:42.36744+01:00"
+ version: 14.3.1
+- name: vespa
+ repository: https://unoplat.github.io/vespa-helm-charts
+ version: 0.2.3
+- name: nginx
+ repository: oci://registry-1.docker.io/bitnamicharts
+ version: 15.14.0
+digest: sha256:ab17b5d2c3883055cb4a26bf530043521be5220c24f804e954bb428273d16ba8
+generated: "2024-05-24T16:55:30.598279-07:00"
diff --git a/deployment/helm/Chart.yaml b/deployment/helm/Chart.yaml
index a36131be126..7763f33bec5 100644
--- a/deployment/helm/Chart.yaml
+++ b/deployment/helm/Chart.yaml
@@ -5,20 +5,31 @@ home: https://www.danswer.ai/
sources:
- "https://github.com/danswer-ai/danswer"
type: application
-version: 0.1.0
-appVersion: "v0.3.42"
+version: 0.2.0
+appVersion: "latest"
annotations:
category: Productivity
licenses: MIT
images: |
- name: webserver
- image: docker.io/danswer/danswer-web-server:v0.3.42
+ image: docker.io/danswer/danswer-web-server:latest
- name: background
- image: docker.io/danswer/danswer-backend:v0.3.42
+ image: docker.io/danswer/danswer-backend:latest
- name: vespa
image: vespaengine/vespa:8.277.17
dependencies:
- name: postgresql
- version: "14.1.0"
+ version: 14.3.1
repository: https://charts.bitnami.com/bitnami
- condition: postgresql.enabled
\ No newline at end of file
+ condition: postgresql.enabled
+ - name: vespa
+ version: 0.2.3
+ repository: https://unoplat.github.io/vespa-helm-charts
+ condition: vespa.enabled
+ - name: nginx
+ version: 15.14.0
+ repository: oci://registry-1.docker.io/bitnamicharts
+ condition: nginx.enabled
+
+
+
\ No newline at end of file
diff --git a/deployment/helm/templates/NOTES.txt b/deployment/helm/templates/NOTES.txt
deleted file mode 100644
index 41703407b6b..00000000000
--- a/deployment/helm/templates/NOTES.txt
+++ /dev/null
@@ -1,22 +0,0 @@
-1. Get the application URL by running these commands:
-{{- if .Values.ingress.enabled }}
-{{- range $host := .Values.ingress.hosts }}
- {{- range .paths }}
- http{{ if $.Values.ingress.tls }}s{{ end }}://{{ $host.host }}{{ .path }}
- {{- end }}
-{{- end }}
-{{- else if contains "NodePort" .Values.webserver.service.type }}
- export NODE_PORT=$(kubectl get --namespace {{ .Release.Namespace }} -o jsonpath="{.spec.ports[0].nodePort}" services {{ include "danswer-stack.fullname" . }})
- export NODE_IP=$(kubectl get nodes --namespace {{ .Release.Namespace }} -o jsonpath="{.items[0].status.addresses[0].address}")
- echo http://$NODE_IP:$NODE_PORT
-{{- else if contains "LoadBalancer" .Values.webserver.service.type }}
- NOTE: It may take a few minutes for the LoadBalancer IP to be available.
- You can watch the status of by running 'kubectl get --namespace {{ .Release.Namespace }} svc -w {{ include "danswer-stack.fullname" . }}'
- export SERVICE_IP=$(kubectl get svc --namespace {{ .Release.Namespace }} {{ include "danswer-stack.fullname" . }} --template "{{"{{ range (index .status.loadBalancer.ingress 0) }}{{.}}{{ end }}"}}")
- echo http://$SERVICE_IP:{{ .Values.webserver.service.port }}
-{{- else if contains "ClusterIP" .Values.webserver.service.type }}
- export POD_NAME=$(kubectl get pods --namespace {{ .Release.Namespace }} -l "app.kubernetes.io/name={{ include "danswer-stack.name" . }},app.kubernetes.io/instance={{ .Release.Name }}" -o jsonpath="{.items[0].metadata.name}")
- export CONTAINER_PORT=$(kubectl get pod --namespace {{ .Release.Namespace }} $POD_NAME -o jsonpath="{.spec.containers[0].ports[0].containerPort}")
- echo "Visit http://127.0.0.1:8080 to use your application"
- kubectl --namespace {{ .Release.Namespace }} port-forward $POD_NAME 8080:$CONTAINER_PORT
-{{- end }}
diff --git a/deployment/helm/templates/_helpers.tpl b/deployment/helm/templates/_helpers.tpl
index 4e6672fd677..483a5b5e5af 100644
--- a/deployment/helm/templates/_helpers.tpl
+++ b/deployment/helm/templates/_helpers.tpl
@@ -60,3 +60,24 @@ Create the name of the service account to use
{{- default "default" .Values.serviceAccount.name }}
{{- end }}
{{- end }}
+
+{{/*
+Set secret name
+*/}}
+{{- define "danswer-stack.secretName" -}}
+{{- default (default "danswer-secrets" .Values.auth.secretName) .Values.auth.existingSecret }}
+{{- end }}
+
+{{/*
+Create env vars from secrets
+*/}}
+{{- define "danswer-stack.envSecrets" -}}
+ {{- range $name, $key := .Values.auth.secretKeys }}
+- name: {{ $name | upper | replace "-" "_" | quote }}
+ valueFrom:
+ secretKeyRef:
+ name: {{ include "danswer-stack.secretName" $ }}
+ key: {{ default $name $key }}
+ {{- end }}
+{{- end }}
+
diff --git a/deployment/helm/templates/api-deployment.yaml b/deployment/helm/templates/api-deployment.yaml
index 8c40f3408c2..7f10bffafd0 100644
--- a/deployment/helm/templates/api-deployment.yaml
+++ b/deployment/helm/templates/api-deployment.yaml
@@ -1,7 +1,7 @@
apiVersion: apps/v1
kind: Deployment
metadata:
- name: {{ include "danswer-stack.fullname" . }}-api
+ name: {{ include "danswer-stack.fullname" . }}-api-deployment
labels:
{{- include "danswer-stack.labels" . | nindent 4 }}
spec:
@@ -11,6 +11,9 @@ spec:
selector:
matchLabels:
{{- include "danswer-stack.selectorLabels" . | nindent 6 }}
+ {{- if .Values.api.deploymentLabels }}
+ {{- toYaml .Values.api.deploymentLabels | nindent 6 }}
+ {{- end }}
template:
metadata:
{{- with .Values.api.podAnnotations }}
@@ -31,7 +34,7 @@ spec:
securityContext:
{{- toYaml .Values.api.podSecurityContext | nindent 8 }}
containers:
- - name: {{ .Chart.Name }}
+ - name: api-server
securityContext:
{{- toYaml .Values.api.securityContext | nindent 12 }}
image: "{{ .Values.api.image.repository }}:{{ .Values.api.image.tag | default .Chart.AppVersion }}"
@@ -51,60 +54,6 @@ spec:
{{- toYaml .Values.api.resources | nindent 12 }}
envFrom:
- configMapRef:
- name: {{ include "danswer-stack.fullname" . }}
+ name: {{ .Values.config.envConfigMapName }}
env:
- - name: INTERNAL_URL
- value: {{ (list "http://" (include "danswer-stack.fullname" .) "-api:" .Values.api.service.port | join "") | quote }}
- - name: VESPA_HOST
- value: {{ (list (include "danswer-stack.fullname" .) "vespa" | join "-") }}
- {{- if .Values.postgresql.enabled }}
- - name: POSTGRES_HOST
- value: {{ (list .Release.Name "postgresql" | join "-") }}
- - name: POSTGRES_DB
- value: {{ .Values.postgresql.auth.database }}
- - name: POSTGRES_USER
- value: {{ .Values.postgresql.auth.username }}
- - name: POSTGRES_PASSWORD
- valueFrom:
- secretKeyRef:
- name: {{ (list .Release.Name "postgresql" | join "-") }}
- key: password
- {{- end }}
- volumeMounts:
- - name: dynamic-storage
- mountPath: /home/storage
- - name: connector-storage
- mountPath: /home/file_connector_storage
- {{- if .Values.api.volumeMounts }}
- {{- .Values.api.volumeMounts | toYaml | nindent 12}}
- {{- end }}
- volumes:
- - name: dynamic-storage
- {{- if .Values.persistence.dynamic.enabled }}
- persistentVolumeClaim:
- claimName: {{ .Values.persistence.dynamic.existingClaim | default (list (include "danswer-stack.fullname" .) "dynamic" | join "-") }}
- {{- else }}
- emptyDir: { }
- {{- end }}
- - name: connector-storage
- {{- if .Values.persistence.connector.enabled }}
- persistentVolumeClaim:
- claimName: {{ .Values.persistence.connector.existingClaim | default (list (include "danswer-stack.fullname" .) "connector" | join "-") }}
- {{- else }}
- emptyDir: { }
- {{- end }}
- {{- if .Values.api.volumes }}
- {{- .Values.api.volumes | toYaml | nindent 8}}
- {{- end }}
- {{- with .Values.api.nodeSelector }}
- nodeSelector:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- {{- with .Values.api.affinity }}
- affinity:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- {{- with .Values.api.tolerations }}
- tolerations:
- {{- toYaml . | nindent 8 }}
- {{- end }}
+ {{- include "danswer-stack.envSecrets" . | nindent 12}}
diff --git a/deployment/helm/templates/api-service.yaml b/deployment/helm/templates/api-service.yaml
index f4e4e0be693..1fd74d4ddf5 100644
--- a/deployment/helm/templates/api-service.yaml
+++ b/deployment/helm/templates/api-service.yaml
@@ -1,9 +1,13 @@
apiVersion: v1
kind: Service
metadata:
- name: {{ include "danswer-stack.fullname" . }}-api
+ # INTERNAL_URL env variable depends on this, don't change without changing INTERNAL_URL
+ name: {{ include "danswer-stack.fullname" . }}-api-service
labels:
{{- include "danswer-stack.labels" . | nindent 4 }}
+ {{- if .Values.api.deploymentLabels }}
+ {{- toYaml .Values.api.deploymentLabels | nindent 4 }}
+ {{- end }}
spec:
type: {{ .Values.api.service.type }}
ports:
@@ -13,3 +17,6 @@ spec:
name: api-server-port
selector:
{{- include "danswer-stack.selectorLabels" . | nindent 4 }}
+ {{- if .Values.api.deploymentLabels }}
+ {{- toYaml .Values.api.deploymentLabels | nindent 4 }}
+ {{- end }}
diff --git a/deployment/helm/templates/background-deployment.yaml b/deployment/helm/templates/background-deployment.yaml
index 59cfc524626..3cd65a99af4 100644
--- a/deployment/helm/templates/background-deployment.yaml
+++ b/deployment/helm/templates/background-deployment.yaml
@@ -11,6 +11,9 @@ spec:
selector:
matchLabels:
{{- include "danswer-stack.selectorLabels" . | nindent 6 }}
+ {{- if .Values.background.deploymentLabels }}
+ {{- toYaml .Values.background.deploymentLabels | nindent 6 }}
+ {{- end }}
template:
metadata:
{{- with .Values.background.podAnnotations }}
@@ -31,7 +34,7 @@ spec:
securityContext:
{{- toYaml .Values.background.podSecurityContext | nindent 8 }}
containers:
- - name: {{ .Chart.Name }}
+ - name: background
securityContext:
{{- toYaml .Values.background.securityContext | nindent 12 }}
image: "{{ .Values.background.image.repository }}:{{ .Values.background.image.tag | default .Chart.AppVersion }}"
@@ -41,60 +44,8 @@ spec:
{{- toYaml .Values.background.resources | nindent 12 }}
envFrom:
- configMapRef:
- name: {{ include "danswer-stack.fullname" . }}
+ name: {{ .Values.config.envConfigMapName }}
env:
- - name: INTERNAL_URL
- value: {{ (list "http://" (include "danswer-stack.fullname" .) "-api:" .Values.api.service.port | join "") | quote }}
- - name: VESPA_HOST
- value: {{ (list (include "danswer-stack.fullname" .) "vespa" | join "-") }}
- {{- if .Values.postgresql.enabled }}
- - name: POSTGRES_HOST
- value: {{ (list .Release.Name "postgresql" | join "-") }}
- - name: POSTGRES_DB
- value: {{ .Values.postgresql.auth.database }}
- - name: POSTGRES_USER
- value: {{ .Values.postgresql.auth.username }}
- - name: POSTGRES_PASSWORD
- valueFrom:
- secretKeyRef:
- name: {{ (list .Release.Name "postgresql" | join "-") }}
- key: password
- {{- end }}
- volumeMounts:
- - name: dynamic-storage
- mountPath: /home/storage
- - name: connector-storage
- mountPath: /home/file_connector_storage
- {{- if .Values.background.volumeMounts }}
- {{- .Values.background.volumeMounts | toYaml | nindent 12}}
- {{- end }}
- volumes:
- - name: dynamic-storage
- {{- if .Values.persistence.dynamic.enabled }}
- persistentVolumeClaim:
- claimName: {{ .Values.persistence.dynamic.existingClaim | default (list (include "danswer-stack.fullname" .) "dynamic" | join "-") }}
- {{- else }}
- emptyDir: { }
- {{- end }}
- - name: connector-storage
- {{- if .Values.persistence.connector.enabled }}
- persistentVolumeClaim:
- claimName: {{ .Values.persistence.connector.existingClaim | default (list (include "danswer-stack.fullname" .) "connector" | join "-") }}
- {{- else }}
- emptyDir: { }
- {{- end }}
- {{- if .Values.background.volumes }}
- {{- .Values.background.volumes | toYaml | nindent 8}}
- {{- end }}
- {{- with .Values.background.nodeSelector }}
- nodeSelector:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- {{- with .Values.background.affinity }}
- affinity:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- {{- with .Values.background.tolerations }}
- tolerations:
- {{- toYaml . | nindent 8 }}
- {{- end }}
+ - name: ENABLE_MINI_CHUNK
+ value: "{{ .Values.background.enableMiniChunk }}"
+ {{- include "danswer-stack.envSecrets" . | nindent 12}}
diff --git a/deployment/helm/templates/configmap.yaml b/deployment/helm/templates/configmap.yaml
index a393977986d..8119ae0459c 100755
--- a/deployment/helm/templates/configmap.yaml
+++ b/deployment/helm/templates/configmap.yaml
@@ -1,11 +1,15 @@
apiVersion: v1
kind: ConfigMap
metadata:
- name: {{ include "danswer-stack.fullname" . }}
+ name: {{ .Values.config.envConfigMapName }}
labels:
{{- include "danswer-stack.labels" . | nindent 4 }}
data:
-{{- range $key, $value := .Values.config }}
- {{ $key }}: |-
- {{- $value | nindent 4 }}
-{{- end }}
+ INTERNAL_URL: "http://{{ include "danswer-stack.fullname" . }}-api-service:{{ .Values.api.service.port | default 8080 }}"
+ POSTGRES_HOST: {{ .Release.Name }}-postgresql
+ VESPA_HOST: "document-index-service"
+ MODEL_SERVER_HOST: "{{ include "danswer-stack.fullname" . }}-inference-model-service"
+ INDEXING_MODEL_SERVER_HOST: "{{ include "danswer-stack.fullname" . }}-indexing-model-service"
+{{- range $key, $value := .Values.configMap }}
+ {{ $key }}: "{{ $value }}"
+{{- end }}
\ No newline at end of file
diff --git a/deployment/helm/templates/connector-pvc.yaml b/deployment/helm/templates/connector-pvc.yaml
deleted file mode 100644
index 41c41c3cffa..00000000000
--- a/deployment/helm/templates/connector-pvc.yaml
+++ /dev/null
@@ -1,19 +0,0 @@
-{{- if and .Values.persistence.connector.enabled (not .Values.persistence.connector.existingClaim)}}
-apiVersion: v1
-kind: PersistentVolumeClaim
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-connector
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- accessModes:
- {{- range .Values.persistence.connector.accessModes }}
- - {{ . | quote }}
- {{- end }}
- resources:
- requests:
- storage: {{ .Values.persistence.connector.size | quote }}
- {{- with .Values.persistence.connector.storageClassName }}
- storageClassName: {{ . }}
- {{- end }}
-{{- end }}
\ No newline at end of file
diff --git a/deployment/helm/templates/danswer-secret.yaml b/deployment/helm/templates/danswer-secret.yaml
new file mode 100644
index 00000000000..6b2aa317204
--- /dev/null
+++ b/deployment/helm/templates/danswer-secret.yaml
@@ -0,0 +1,11 @@
+{{- if not .Values.auth.existingSecret -}}
+apiVersion: v1
+kind: Secret
+metadata:
+ name: {{ include "danswer-stack.secretName" . }}
+type: Opaque
+stringData:
+ {{- range $name, $value := .Values.auth.secrets }}
+ {{ $name }}: {{ $value | quote }}
+ {{- end }}
+{{- end }}
\ No newline at end of file
diff --git a/deployment/helm/templates/dynamic-pvc.yaml b/deployment/helm/templates/dynamic-pvc.yaml
deleted file mode 100644
index 703b33acb59..00000000000
--- a/deployment/helm/templates/dynamic-pvc.yaml
+++ /dev/null
@@ -1,19 +0,0 @@
-{{- if and .Values.persistence.dynamic.enabled (not .Values.persistence.dynamic.existingClaim)}}
-apiVersion: v1
-kind: PersistentVolumeClaim
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-dynamic
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- accessModes:
- {{- range .Values.persistence.dynamic.accessModes }}
- - {{ . | quote }}
- {{- end }}
- resources:
- requests:
- storage: {{ .Values.persistence.dynamic.size | quote }}
- {{- with .Values.persistence.dynamic.storageClassName }}
- storageClassName: {{ . }}
- {{- end }}
-{{- end }}
\ No newline at end of file
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/indexing-model-deployment.yaml b/deployment/helm/templates/indexing-model-deployment.yaml
similarity index 91%
rename from deployment/kubernetes/charts/danswer-stack/templates/indexing-model-deployment.yaml
rename to deployment/helm/templates/indexing-model-deployment.yaml
index 5bd866b55c4..cc88aefb79a 100644
--- a/deployment/kubernetes/charts/danswer-stack/templates/indexing-model-deployment.yaml
+++ b/deployment/helm/templates/indexing-model-deployment.yaml
@@ -34,11 +34,10 @@ spec:
envFrom:
- configMapRef:
name: {{ .Values.config.envConfigMapName }}
- {{- if .Values.indexCapability.indexingOnly }}
env:
- name: INDEXING_ONLY
- value: "{{ .Values.indexCapability.indexingOnly }}"
- {{- end }}
+ value: "{{ default "True" .Values.indexCapability.indexingOnly }}"
+ {{- include "danswer-stack.envSecrets" . | nindent 10}}
volumeMounts:
{{- range .Values.indexCapability.volumeMounts }}
- name: {{ .name }}
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/indexing-model-pvc.yaml b/deployment/helm/templates/indexing-model-pvc.yaml
similarity index 100%
rename from deployment/kubernetes/charts/danswer-stack/templates/indexing-model-pvc.yaml
rename to deployment/helm/templates/indexing-model-pvc.yaml
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/indexing-model-service.yaml b/deployment/helm/templates/indexing-model-service.yaml
similarity index 100%
rename from deployment/kubernetes/charts/danswer-stack/templates/indexing-model-service.yaml
rename to deployment/helm/templates/indexing-model-service.yaml
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/inference-model-deployment.yaml b/deployment/helm/templates/inference-model-deployment.yaml
similarity index 95%
rename from deployment/kubernetes/charts/danswer-stack/templates/inference-model-deployment.yaml
rename to deployment/helm/templates/inference-model-deployment.yaml
index a09dfa4fbb5..43caddd29c3 100644
--- a/deployment/kubernetes/charts/danswer-stack/templates/inference-model-deployment.yaml
+++ b/deployment/helm/templates/inference-model-deployment.yaml
@@ -30,6 +30,8 @@ spec:
envFrom:
- configMapRef:
name: {{ .Values.config.envConfigMapName }}
+ env:
+ {{- include "danswer-stack.envSecrets" . | nindent 12}}
volumeMounts:
{{- range .Values.inferenceCapability.deployment.volumeMounts }}
- name: {{ .name }}
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/inference-model-pvc.yaml b/deployment/helm/templates/inference-model-pvc.yaml
similarity index 100%
rename from deployment/kubernetes/charts/danswer-stack/templates/inference-model-pvc.yaml
rename to deployment/helm/templates/inference-model-pvc.yaml
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/inference-model-service.yaml b/deployment/helm/templates/inference-model-service.yaml
similarity index 100%
rename from deployment/kubernetes/charts/danswer-stack/templates/inference-model-service.yaml
rename to deployment/helm/templates/inference-model-service.yaml
diff --git a/deployment/helm/templates/ingress.yaml b/deployment/helm/templates/ingress.yaml
deleted file mode 100644
index cfbef35dd7d..00000000000
--- a/deployment/helm/templates/ingress.yaml
+++ /dev/null
@@ -1,60 +0,0 @@
-{{- if .Values.ingress.enabled -}}
-{{- $fullName := include "danswer-stack.fullname" . -}}
-{{- if and .Values.ingress.className (not (semverCompare ">=1.18-0" .Capabilities.KubeVersion.GitVersion)) }}
- {{- if not (hasKey .Values.ingress.annotations "kubernetes.io/ingress.class") }}
- {{- $_ := set .Values.ingress.annotations "kubernetes.io/ingress.class" .Values.ingress.className}}
- {{- end }}
-{{- end }}
-{{- if semverCompare ">=1.19-0" .Capabilities.KubeVersion.GitVersion -}}
-apiVersion: networking.k8s.io/v1
-{{- else if semverCompare ">=1.14-0" .Capabilities.KubeVersion.GitVersion -}}
-apiVersion: networking.k8s.io/v1beta1
-{{- else -}}
-apiVersion: extensions/v1beta1
-{{- end }}
-kind: Ingress
-metadata:
- name: {{ $fullName }}
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
- {{- with .Values.ingress.annotations }}
- annotations:
- {{- toYaml . | nindent 4 }}
- {{- end }}
-spec:
- {{- if and .Values.ingress.className (semverCompare ">=1.18-0" .Capabilities.KubeVersion.GitVersion) }}
- ingressClassName: {{ .Values.ingress.className }}
- {{- end }}
- {{- if .Values.ingress.tls }}
- tls:
- {{- range .Values.ingress.tls }}
- - hosts:
- {{- range .hosts }}
- - {{ . | quote }}
- {{- end }}
- secretName: {{ .secretName }}
- {{- end }}
- {{- end }}
- rules:
- {{- range .Values.ingress.hosts }}
- - host: {{ .host | quote }}
- http:
- paths:
- {{- range .paths }}
- - path: {{ .path }}
- {{- if and .pathType (semverCompare ">=1.18-0" $.Capabilities.KubeVersion.GitVersion) }}
- pathType: {{ .pathType }}
- {{- end }}
- backend:
- {{- if semverCompare ">=1.19-0" $.Capabilities.KubeVersion.GitVersion }}
- service:
- name: {{ (list $fullName .service) | join "-" }}
- port:
- number: {{ .servicePort }}
- {{- else }}
- serviceName: {{ (list $fullName .service) | join "-" }}
- servicePort: {{ .servicePort }}
- {{- end }}
- {{- end }}
- {{- end }}
-{{- end }}
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/nginx-conf.yaml b/deployment/helm/templates/nginx-conf.yaml
similarity index 92%
rename from deployment/kubernetes/charts/danswer-stack/templates/nginx-conf.yaml
rename to deployment/helm/templates/nginx-conf.yaml
index 383151ea1d3..81ecbaaa2f6 100644
--- a/deployment/kubernetes/charts/danswer-stack/templates/nginx-conf.yaml
+++ b/deployment/helm/templates/nginx-conf.yaml
@@ -5,7 +5,7 @@ metadata:
data:
nginx.conf: |
upstream api_server {
- server {{ include "danswer-stack.fullname" . }}-api:{{ .Values.api.service.port }} fail_timeout=0;
+ server {{ include "danswer-stack.fullname" . }}-api-service:{{ .Values.api.service.port }} fail_timeout=0;
}
upstream web_server {
diff --git a/deployment/helm/templates/secret.yaml b/deployment/helm/templates/secret.yaml
deleted file mode 100755
index 58bfba87d95..00000000000
--- a/deployment/helm/templates/secret.yaml
+++ /dev/null
@@ -1,10 +0,0 @@
-apiVersion: v1
-kind: Secret
-metadata:
- name: {{ include "danswer-stack.fullname" . }}
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-data:
-{{- range $key, $value := .Values.secrets }}
- {{ $key }}: '{{ $value | b64enc }}'
-{{- end }}
diff --git a/deployment/helm/templates/vespa-service.yaml b/deployment/helm/templates/vespa-service.yaml
deleted file mode 100644
index 01216a28970..00000000000
--- a/deployment/helm/templates/vespa-service.yaml
+++ /dev/null
@@ -1,23 +0,0 @@
-apiVersion: v1
-kind: Service
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-vespa
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- type: ClusterIP
- ports:
- - name: vespa-tenant-port
- protocol: TCP
- port: 19070
- targetPort: 19070
- - name: vespa-tenant-port-2
- protocol: TCP
- port: 19071
- targetPort: 19071
- - name: vespa-port
- protocol: TCP
- port: 8080
- targetPort: 8080
- selector:
- {{- include "danswer-stack.selectorLabels" . | nindent 4 }}
diff --git a/deployment/helm/templates/vespa-statefulset.yaml b/deployment/helm/templates/vespa-statefulset.yaml
deleted file mode 100644
index 674b52bc447..00000000000
--- a/deployment/helm/templates/vespa-statefulset.yaml
+++ /dev/null
@@ -1,83 +0,0 @@
-apiVersion: apps/v1
-kind: StatefulSet
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-vespa
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- replicas: {{ .Values.vespa.replicaCount }}
- selector:
- matchLabels:
- {{- include "danswer-stack.selectorLabels" . | nindent 6 }}
- template:
- metadata:
- {{- with .Values.vespa.podAnnotations }}
- annotations:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- labels:
- {{- include "danswer-stack.labels" . | nindent 8 }}
- {{- with .Values.vespa.podLabels }}
- {{- toYaml . | nindent 8 }}
- {{- end }}
- spec:
- {{- with .Values.imagePullSecrets }}
- imagePullSecrets:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- serviceAccountName: {{ include "danswer-stack.serviceAccountName" . }}
- securityContext:
- {{- toYaml .Values.vespa.podSecurityContext | nindent 8 }}
- containers:
- - name: {{ .Chart.Name }}
- securityContext:
- {{- toYaml .Values.vespa.securityContext | nindent 12 }}
- image: "{{ .Values.vespa.image.repository }}:{{ .Values.vespa.image.tag }}"
- imagePullPolicy: {{ .Values.vespa.image.pullPolicy }}
- ports:
- - containerPort: 19070
- - containerPort: 19071
- - containerPort: 8081
- livenessProbe:
- httpGet:
- path: /state/v1/health
- port: 19071
- scheme: HTTP
- readinessProbe:
- httpGet:
- path: /state/v1/health
- port: 19071
- scheme: HTTP
- resources:
- {{- toYaml .Values.vespa.resources | nindent 12 }}
- volumeMounts:
- - name: vespa-storage
- mountPath: /opt/vespa/var/
- {{- with .Values.vespa.nodeSelector }}
- nodeSelector:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- {{- with .Values.vespa.affinity }}
- affinity:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- {{- with .Values.vespa.tolerations }}
- tolerations:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- {{- if .Values.persistence.vespa.enabled }}
- volumeClaimTemplates:
- - metadata:
- name: vespa-storage
- spec:
- accessModes:
- {{- range .Values.persistence.vespa.accessModes }}
- - {{ . | quote }}
- {{- end }}
- resources:
- requests:
- storage: {{ .Values.persistence.vespa.size | quote }}
- {{- with .Values.persistence.vespa.storageClassName }}
- storageClassName: {{ . }}
- {{- end }}
- {{- end }}
\ No newline at end of file
diff --git a/deployment/helm/templates/webserver-deployment.yaml b/deployment/helm/templates/webserver-deployment.yaml
index c679e6e0a21..c3505248fc6 100644
--- a/deployment/helm/templates/webserver-deployment.yaml
+++ b/deployment/helm/templates/webserver-deployment.yaml
@@ -11,6 +11,9 @@ spec:
selector:
matchLabels:
{{- include "danswer-stack.selectorLabels" . | nindent 6 }}
+ {{- if .Values.webserver.deploymentLabels }}
+ {{- toYaml .Values.webserver.deploymentLabels | nindent 6 }}
+ {{- end }}
template:
metadata:
{{- with .Values.webserver.podAnnotations }}
@@ -31,7 +34,7 @@ spec:
securityContext:
{{- toYaml .Values.webserver.podSecurityContext | nindent 8 }}
containers:
- - name: {{ .Chart.Name }}
+ - name: web-server
securityContext:
{{- toYaml .Values.webserver.securityContext | nindent 12 }}
image: "{{ .Values.webserver.image.repository }}:{{ .Values.webserver.image.tag | default .Chart.AppVersion }}"
@@ -40,37 +43,13 @@ spec:
- name: http
containerPort: {{ .Values.webserver.service.port }}
protocol: TCP
- livenessProbe:
- httpGet:
- path: /
- port: http
- readinessProbe:
- httpGet:
- path: /
- port: http
resources:
{{- toYaml .Values.webserver.resources | nindent 12 }}
envFrom:
- configMapRef:
- name: {{ include "danswer-stack.fullname" . }}
+ name: {{ .Values.config.envConfigMapName }}
env:
- - name: INTERNAL_URL
- value: {{ (list "http://" (include "danswer-stack.fullname" .) "-api:" .Values.api.service.port | join "") | quote }}
- - name: VESPA_HOST
- value: {{ (list (include "danswer-stack.fullname" .) "vespa" | join "-") }}
- {{- if .Values.postgresql.enabled }}
- - name: POSTGRES_HOST
- value: {{ (list .Release.Name "postgresql" | join "-") }}
- - name: POSTGRES_DB
- value: {{ .Values.postgresql.auth.database }}
- - name: POSTGRES_USER
- value: {{ .Values.postgresql.auth.username }}
- - name: POSTGRES_PASSWORD
- valueFrom:
- secretKeyRef:
- name: {{ (list .Release.Name "postgresql" | join "-") }}
- key: password
- {{- end }}
+ {{- include "danswer-stack.envSecrets" . | nindent 12}}
{{- with .Values.webserver.volumeMounts }}
volumeMounts:
{{- toYaml . | nindent 12 }}
@@ -79,15 +58,3 @@ spec:
volumes:
{{- toYaml . | nindent 8 }}
{{- end }}
- {{- with .Values.webserver.nodeSelector }}
- nodeSelector:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- {{- with .Values.webserver.affinity }}
- affinity:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- {{- with .Values.webserver.tolerations }}
- tolerations:
- {{- toYaml . | nindent 8 }}
- {{- end }}
diff --git a/deployment/helm/templates/webserver-service.yaml b/deployment/helm/templates/webserver-service.yaml
index 776b65f8f96..3e33566fce1 100644
--- a/deployment/helm/templates/webserver-service.yaml
+++ b/deployment/helm/templates/webserver-service.yaml
@@ -4,6 +4,9 @@ metadata:
name: {{ include "danswer-stack.fullname" . }}-webserver
labels:
{{- include "danswer-stack.labels" . | nindent 4 }}
+ {{- if .Values.webserver.deploymentLabels }}
+ {{- toYaml .Values.webserver.deploymentLabels | nindent 4 }}
+ {{- end }}
spec:
type: {{ .Values.webserver.service.type }}
ports:
@@ -13,3 +16,6 @@ spec:
name: http
selector:
{{- include "danswer-stack.selectorLabels" . | nindent 4 }}
+ {{- if .Values.webserver.deploymentLabels }}
+ {{- toYaml .Values.webserver.deploymentLabels | nindent 4 }}
+ {{- end }}
diff --git a/deployment/helm/values.yaml b/deployment/helm/values.yaml
index 8d994b55ff3..8ef4167a6eb 100644
--- a/deployment/helm/values.yaml
+++ b/deployment/helm/values.yaml
@@ -6,9 +6,68 @@ imagePullSecrets: []
nameOverride: ""
fullnameOverride: ""
+inferenceCapability:
+ service:
+ name: inference-model-server-service
+ type: ClusterIP
+ port: 9000
+ pvc:
+ name: inference-model-pvc
+ accessModes:
+ - ReadWriteOnce
+ storage: 3Gi
+ deployment:
+ name: inference-model-server-deployment
+ replicas: 1
+ labels:
+ - key: app
+ value: inference-model-server
+ image:
+ repository: danswer/danswer-model-server
+ tag: latest
+ pullPolicy: IfNotPresent
+ command: ["uvicorn", "model_server.main:app", "--host", "0.0.0.0", "--port", "9000"]
+ port: 9000
+ volumeMounts:
+ - name: inference-model-storage
+ mountPath: /root/.cache
+ volumes:
+ - name: inference-model-storage
+ persistentVolumeClaim:
+ claimName: inference-model-pvc
+ podLabels:
+ - key: app
+ value: inference-model-server
+
+indexCapability:
+ service:
+ type: ClusterIP
+ port: 9000
+ name: indexing-model-server-port
+ deploymentLabels:
+ app: indexing-model-server
+ podLabels:
+ app: indexing-model-server
+ indexingOnly: "True"
+ podAnnotations: {}
+ volumeMounts:
+ - name: indexing-model-storage
+ mountPath: /root/.cache
+ volumes:
+ - name: indexing-model-storage
+ persistentVolumeClaim:
+ claimName: indexing-model-storage
+ indexingModelPVC:
+ name: indexing-model-storage
+ accessMode: "ReadWriteOnce"
+ storage: "3Gi"
+
+config:
+ envConfigMapName: env-configmap
+
serviceAccount:
# Specifies whether a service account should be created
- create: true
+ create: false
# Automatically mount a ServiceAccount's API credentials?
automount: true
# Annotations to add to the service account
@@ -17,6 +76,31 @@ serviceAccount:
# If not set and create is true, a name is generated using the fullname template
name: ""
+postgresql:
+ primary:
+ persistence:
+ size: 5Gi
+ enabled: true
+ auth:
+ existingSecret: danswer-secrets
+ secretKeys:
+ adminPasswordKey: postgres_password #overwriting as postgres typically expects 'postgres-password'
+
+nginx:
+ containerPorts:
+ http: 1024
+ extraEnvVars:
+ - name: DOMAIN
+ value: localhost
+ service:
+ ports:
+ http: 80
+ danswer: 3000
+ targetPort:
+ http: http
+ danswer: http
+
+ existingServerBlockConfigmap: danswer-nginx-conf
webserver:
replicaCount: 1
@@ -25,10 +109,11 @@ webserver:
pullPolicy: IfNotPresent
# Overrides the image tag whose default is the chart appVersion.
tag: ""
-
+ deploymentLabels:
+ app: web-server
podAnnotations: {}
- podLabels: {}
-
+ podLabels:
+ app: web-server
podSecurityContext: {}
# fsGroup: 2000
@@ -87,10 +172,12 @@ api:
pullPolicy: IfNotPresent
# Overrides the image tag whose default is the chart appVersion.
tag: ""
-
+ deploymentLabels:
+ app: api-server
podAnnotations: {}
podLabels:
scope: danswer-backend
+ app: api-server
podSecurityContext: {}
# fsGroup: 2000
@@ -107,17 +194,17 @@ api:
type: ClusterIP
port: 8080
- resources:
+ resources: {}
# We usually recommend not to specify default resources and to leave this as a conscious
# choice for the user. This also increases chances charts run on environments with little
# resources, such as Minikube. If you do want to specify resources, uncomment the following
# lines, adjust them as necessary, and remove the curly braces after 'resources:'.
- requests:
- cpu: 1500m
- memory: 2Gi
- # limits:
- # cpu: 100m
- # memory: 128Mi
+ # requests:
+ # cpu: 1000m # Requests 1 CPU core
+ # memory: 1Gi # Requests 1 GiB of memory
+ # limits:
+ # cpu: 2000m # Limits to 2 CPU cores
+ # memory: 2Gi # Limits to 2 GiB of memory
autoscaling:
enabled: false
@@ -141,16 +228,7 @@ api:
nodeSelector: {}
tolerations: []
- affinity:
- podAffinity:
- requiredDuringSchedulingIgnoredDuringExecution:
- - labelSelector:
- matchExpressions:
- - key: scope
- operator: In
- values:
- - danswer-backend
- topologyKey: "kubernetes.io/hostname"
+
background:
replicaCount: 1
@@ -158,11 +236,13 @@ background:
repository: danswer/danswer-backend
pullPolicy: IfNotPresent
# Overrides the image tag whose default is the chart appVersion.
- tag: ""
+ tag: latest
podAnnotations: {}
podLabels:
scope: danswer-backend
-
+ app: background
+ deploymentLabels:
+ app: background
podSecurityContext: {}
# fsGroup: 2000
@@ -173,18 +253,18 @@ background:
# readOnlyRootFilesystem: true
# runAsNonRoot: true
# runAsUser: 1000
-
- resources:
+ enableMiniChunk: "true"
+ resources: {}
# We usually recommend not to specify default resources and to leave this as a conscious
# choice for the user. This also increases chances charts run on environments with little
# resources, such as Minikube. If you do want to specify resources, uncomment the following
# lines, adjust them as necessary, and remove the curly braces after 'resources:'.
- requests:
- cpu: 2500m
- memory: 5Gi
- # limits:
- # cpu: 100m
- # memory: 128Mi
+ # requests:
+ # cpu: 1000m # Requests 1 CPU core
+ # memory: 1Gi # Requests 1 GiB of memory
+ # limits:
+ # cpu: 2000m # Limits to 2 CPU cores
+ # memory: 2Gi # Limits to 2 GiB of memory
autoscaling:
enabled: false
@@ -208,25 +288,19 @@ background:
nodeSelector: {}
tolerations: []
- affinity:
- podAffinity:
- requiredDuringSchedulingIgnoredDuringExecution:
- - labelSelector:
- matchExpressions:
- - key: scope
- operator: In
- values:
- - danswer-backend
- topologyKey: "kubernetes.io/hostname"
vespa:
replicaCount: 1
image:
- repository: vespaengine/vespa
+ repository: vespa
pullPolicy: IfNotPresent
tag: "8.277.17"
podAnnotations: {}
- podLabels: {}
+ podLabels:
+ app: vespa
+ app.kubernetes.io/instance: danswer
+ app.kubernetes.io/name: vespa
+ enabled: true
podSecurityContext: {}
# fsGroup: 2000
@@ -242,16 +316,14 @@ vespa:
# runAsUser: 1000
resources:
- # We usually recommend not to specify default resources and to leave this as a conscious
- # choice for the user. This also increases chances charts run on environments with little
- # resources, such as Minikube. If you do want to specify resources, uncomment the following
- # lines, adjust them as necessary, and remove the curly braces after 'resources:'.
+ # The Vespa Helm chart specifies default resources, which are quite modest. We override
+ # them here to increase chances of the chart running successfully.
requests:
- cpu: 2500m
- memory: 5Gi
- # limits:
- # cpu: 100m
- # memory: 128Mi
+ cpu: 1500m
+ memory: 4000Mi
+ # limits:
+ # cpu: 100m
+ # memory: 128Mi
nodeSelector: {}
tolerations: []
@@ -281,97 +353,99 @@ persistence:
storageClassName: ""
accessModes:
- ReadWriteOnce
- size: 1Gi
- connector:
- enabled: true
- existingClaim: ""
- storageClassName: ""
- accessModes:
- - ReadWriteOnce
- size: 1Gi
- dynamic:
- enabled: true
- existingClaim: ""
- storageClassName: ""
- accessModes:
- - ReadWriteOnce
- size: 1Gi
-
-postgresql:
- enabled: false
- auth:
- postgresPassword: ""
- username: danswer
- password: danswer
- database: danswer
-
-config:
- # Auth Setting, also check the secrets file
- #AUTH_TYPE: "disabled" # Change this for production uses unless Danswer is only accessible behind VPN
- #SESSION_EXPIRE_TIME_SECONDS: "86400" # 1 Day Default
- #VALID_EMAIL_DOMAINS: "" # Can be something like danswer.ai, as an extra double-check
- #SMTP_SERVER: "" # For sending verification emails, if unspecified then defaults to 'smtp.gmail.com'
- #SMTP_PORT: "" # For sending verification emails, if unspecified then defaults to '587'
- #SMTP_USER: "" # 'your-email@company.com'
- #SMTP_PASS: "" # 'your-gmail-password'
- #EMAIL_FROM: "" # 'your-email@company.com' SMTP_USER missing used instead
+ size: 5Gi
+
+auth:
+ # for storing smtp, oauth, slack, and other secrets
+ # keys are lowercased version of env vars (e.g. SMTP_USER -> smtp_user)
+ existingSecret: "" # danswer-secrets
+ # optionally override the secret keys to reference in the secret
+ secretKeys:
+ postgres_password: "postgres_password"
+ smtp_pass: ""
+ oauth_client_id: ""
+ oauth_client_secret: ""
+ oauth_cookie_secret: ""
+ gen_ai_api_key: ""
+ danswer_bot_slack_app_token: ""
+ danswer_bot_slack_bot_token: ""
+ # will be overridden by the existingSecret if set
+ secretName: "danswer-secrets"
+ # set values as strings, they will be base64 encoded
+ secrets:
+ postgres_password: "postgres"
+ smtp_pass: ""
+ oauth_client_id: ""
+ oauth_client_secret: ""
+ oauth_cookie_secret: ""
+ gen_ai_api_key: ""
+ danswer_bot_slack_app_token: ""
+ danswer_bot_slack_bot_token: ""
+
+configMap:
+ AUTH_TYPE: "disabled" # Change this for production uses unless Danswer is only accessible behind VPN
+ SESSION_EXPIRE_TIME_SECONDS: "86400" # 1 Day Default
+ VALID_EMAIL_DOMAINS: "" # Can be something like danswer.ai, as an extra double-check
+ SMTP_SERVER: "" # For sending verification emails, if unspecified then defaults to 'smtp.gmail.com'
+ SMTP_PORT: "" # For sending verification emails, if unspecified then defaults to '587'
+ SMTP_USER: "" # 'your-email@company.com'
+ # SMTP_PASS: "" # 'your-gmail-password'
+ EMAIL_FROM: "" # 'your-email@company.com' SMTP_USER missing used instead
# Gen AI Settings
- #GEN_AI_MODEL_PROVIDER: "openai"
- #GEN_AI_MODEL_VERSION: "gpt-4" # "gpt-3.5-turbo-0125" # Use GPT-4 if you have it
- #FAST_GEN_AI_MODEL_VERSION: "gpt-3.5-turbo-0125"
- #GEN_AI_API_KEY: ""
- #GEN_AI_API_ENDPOINT: ""
- #GEN_AI_API_VERSION: ""
- #GEN_AI_LLM_PROVIDER_TYPE: ""
- #GEN_AI_MAX_TOKENS: ""
- #QA_TIMEOUT: "60"
- #MAX_CHUNKS_FED_TO_CHAT: ""
- #DISABLE_LLM_FILTER_EXTRACTION: ""
- #DISABLE_LLM_CHUNK_FILTER: ""
- #DISABLE_LLM_CHOOSE_SEARCH: ""
+ GEN_AI_MODEL_PROVIDER: ""
+ GEN_AI_MODEL_VERSION: ""
+ FAST_GEN_AI_MODEL_VERSION: ""
+ # GEN_AI_API_KEY: ""
+ GEN_AI_API_ENDPOINT: ""
+ GEN_AI_API_VERSION: ""
+ GEN_AI_LLM_PROVIDER_TYPE: ""
+ GEN_AI_MAX_TOKENS: ""
+ QA_TIMEOUT: "60"
+ MAX_CHUNKS_FED_TO_CHAT: ""
+ DISABLE_LLM_FILTER_EXTRACTION: ""
+ DISABLE_LLM_CHUNK_FILTER: ""
+ DISABLE_LLM_CHOOSE_SEARCH: ""
+ DISABLE_LLM_QUERY_REPHRASE: ""
# Query Options
- #DOC_TIME_DECAY: ""
- #HYBRID_ALPHA: ""
- #EDIT_KEYWORD_QUERY: ""
- #MULTILINGUAL_QUERY_EXPANSION: ""
- #QA_PROMPT_OVERRIDE: ""
+ DOC_TIME_DECAY: ""
+ HYBRID_ALPHA: ""
+ EDIT_KEYWORD_QUERY: ""
+ MULTILINGUAL_QUERY_EXPANSION: ""
+ QA_PROMPT_OVERRIDE: ""
# Don't change the NLP models unless you know what you're doing
- #DOCUMENT_ENCODER_MODEL: ""
- #NORMALIZE_EMBEDDINGS: ""
- #ASYM_QUERY_PREFIX: ""
- #ASYM_PASSAGE_PREFIX: ""
- #ENABLE_RERANKING_REAL_TIME_FLOW: ""
- #ENABLE_RERANKING_ASYNC_FLOW: ""
- #MODEL_SERVER_HOST: ""
- #MODEL_SERVER_PORT: ""
- #INDEXING_MODEL_SERVER_HOST: ""
- #MIN_THREADS_ML_MODELS: ""
+ DOCUMENT_ENCODER_MODEL: ""
+ NORMALIZE_EMBEDDINGS: ""
+ ASYM_QUERY_PREFIX: ""
+ ASYM_PASSAGE_PREFIX: ""
+ ENABLE_RERANKING_REAL_TIME_FLOW: ""
+ ENABLE_RERANKING_ASYNC_FLOW: ""
+ MODEL_SERVER_PORT: ""
+ MIN_THREADS_ML_MODELS: ""
# Indexing Configs
- #NUM_INDEXING_WORKERS: ""
- #DASK_JOB_CLIENT_ENABLED: ""
- #CONTINUE_ON_CONNECTOR_FAILURE: ""
- #EXPERIMENTAL_CHECKPOINTING_ENABLED: ""
- #CONFLUENCE_CONNECTOR_LABELS_TO_SKIP: ""
- #GONG_CONNECTOR_START_TIME: ""
- #NOTION_CONNECTOR_ENABLE_RECURSIVE_PAGE_LOOKUP: ""
+ NUM_INDEXING_WORKERS: ""
+ DISABLE_INDEX_UPDATE_ON_SWAP: ""
+ DASK_JOB_CLIENT_ENABLED: ""
+ CONTINUE_ON_CONNECTOR_FAILURE: ""
+ EXPERIMENTAL_CHECKPOINTING_ENABLED: ""
+ CONFLUENCE_CONNECTOR_LABELS_TO_SKIP: ""
+ JIRA_API_VERSION: ""
+ GONG_CONNECTOR_START_TIME: ""
+ NOTION_CONNECTOR_ENABLE_RECURSIVE_PAGE_LOOKUP: ""
# DanswerBot SlackBot Configs
- #DANSWER_BOT_SLACK_APP_TOKEN: ""
- #DANSWER_BOT_SLACK_BOT_TOKEN: ""
- #DANSWER_BOT_DISABLE_DOCS_ONLY_ANSWER: ""
- #DANSWER_BOT_DISPLAY_ERROR_MSGS: ""
- #DANSWER_BOT_RESPOND_EVERY_CHANNEL: ""
- #DANSWER_BOT_DISABLE_COT: "" # Currently unused
- #NOTIFY_SLACKBOT_NO_ANSWER: ""
+ # DANSWER_BOT_SLACK_APP_TOKEN: ""
+ # DANSWER_BOT_SLACK_BOT_TOKEN: ""
+ DANSWER_BOT_DISABLE_DOCS_ONLY_ANSWER: ""
+ DANSWER_BOT_DISPLAY_ERROR_MSGS: ""
+ DANSWER_BOT_RESPOND_EVERY_CHANNEL: ""
+ DANSWER_BOT_DISABLE_COT: "" # Currently unused
+ NOTIFY_SLACKBOT_NO_ANSWER: ""
# Logging
# Optional Telemetry, please keep it on (nothing sensitive is collected)? <3
# https://docs.danswer.dev/more/telemetry
- #DISABLE_TELEMETRY: ""
- #LOG_LEVEL: ""
- #LOG_ALL_MODEL_INTERACTIONS: ""
- #LOG_VESPA_TIMING_INFORMATION: ""
+ DISABLE_TELEMETRY: ""
+ LOG_LEVEL: ""
+ LOG_ALL_MODEL_INTERACTIONS: ""
+ LOG_VESPA_TIMING_INFORMATION: ""
# Shared or Non-backend Related
- #INTERNAL_URL: "http://api-server-service:80" # for web server
WEB_DOMAIN: "http://localhost:3000" # for web server and api server
- # Other Services
- #POSTGRES_HOST: "relational-db-service"
- #VESPA_HOST: "document-index-service"
\ No newline at end of file
+ DOMAIN: "localhost" # for nginx
diff --git a/deployment/kubernetes/api_server-service-deployment.yaml b/deployment/kubernetes/api_server-service-deployment.yaml
index 63d86ded5eb..eeac5fecc96 100644
--- a/deployment/kubernetes/api_server-service-deployment.yaml
+++ b/deployment/kubernetes/api_server-service-deployment.yaml
@@ -41,18 +41,17 @@ spec:
- containerPort: 8080
# There are some extra values since this is shared between services
# There are no conflicts though, extra env variables are simply ignored
+ env:
+ - name: OAUTH_CLIENT_ID
+ valueFrom:
+ secretKeyRef:
+ name: danswer-secrets
+ key: google_oauth_client_id
+ - name: OAUTH_CLIENT_SECRET
+ valueFrom:
+ secretKeyRef:
+ name: danswer-secrets
+ key: google_oauth_client_secret
envFrom:
- configMapRef:
name: env-configmap
- volumeMounts:
- - name: dynamic-storage
- mountPath: /home/storage
- - name: file-connector-storage
- mountPath: /home/file_connector_storage
- volumes:
- - name: dynamic-storage
- persistentVolumeClaim:
- claimName: dynamic-pvc
- - name: file-connector-storage
- persistentVolumeClaim:
- claimName: file-connector-pvc
diff --git a/deployment/kubernetes/background-deployment.yaml b/deployment/kubernetes/background-deployment.yaml
index 77bfc65fe09..82369e6d3e8 100644
--- a/deployment/kubernetes/background-deployment.yaml
+++ b/deployment/kubernetes/background-deployment.yaml
@@ -22,15 +22,3 @@ spec:
envFrom:
- configMapRef:
name: env-configmap
- volumeMounts:
- - name: dynamic-storage
- mountPath: /home/storage
- - name: file-connector-storage
- mountPath: /home/file_connector_storage
- volumes:
- - name: dynamic-storage
- persistentVolumeClaim:
- claimName: dynamic-pvc
- - name: file-connector-storage
- persistentVolumeClaim:
- claimName: file-connector-pvc
diff --git a/deployment/kubernetes/charts/danswer-stack/.gitignore b/deployment/kubernetes/charts/danswer-stack/.gitignore
deleted file mode 100644
index b442275d6b5..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/.gitignore
+++ /dev/null
@@ -1,3 +0,0 @@
-### Helm ###
-# Chart dependencies
-**/charts/*.tgz
diff --git a/deployment/kubernetes/charts/danswer-stack/.helmignore b/deployment/kubernetes/charts/danswer-stack/.helmignore
deleted file mode 100644
index 0e8a0eb36f4..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/.helmignore
+++ /dev/null
@@ -1,23 +0,0 @@
-# Patterns to ignore when building packages.
-# This supports shell glob matching, relative path matching, and
-# negation (prefixed with !). Only one pattern per line.
-.DS_Store
-# Common VCS dirs
-.git/
-.gitignore
-.bzr/
-.bzrignore
-.hg/
-.hgignore
-.svn/
-# Common backup files
-*.swp
-*.bak
-*.tmp
-*.orig
-*~
-# Various IDEs
-.project
-.idea/
-*.tmproj
-.vscode/
diff --git a/deployment/kubernetes/charts/danswer-stack/Chart.lock b/deployment/kubernetes/charts/danswer-stack/Chart.lock
deleted file mode 100644
index c7f54c8c1b9..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/Chart.lock
+++ /dev/null
@@ -1,12 +0,0 @@
-dependencies:
-- name: postgresql
- repository: https://charts.bitnami.com/bitnami
- version: 14.3.1
-- name: vespa
- repository: https://unoplat.github.io/vespa-helm-charts
- version: 0.2.2
-- name: nginx
- repository: oci://registry-1.docker.io/bitnamicharts
- version: 15.14.0
-digest: sha256:53e138c0ab12193f57a76c2f377e2a5d3d11c394b03eef5f6848dfae6705cb61
-generated: "2024-03-27T12:34:11.548396+05:30"
diff --git a/deployment/kubernetes/charts/danswer-stack/Chart.yaml b/deployment/kubernetes/charts/danswer-stack/Chart.yaml
deleted file mode 100644
index 0819afa8725..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/Chart.yaml
+++ /dev/null
@@ -1,35 +0,0 @@
-apiVersion: v2
-name: danswer-stack
-description: A Helm chart for Kubernetes
-home: https://www.danswer.ai/
-sources:
- - "https://github.com/danswer-ai/danswer"
-type: application
-version: 0.1.0
-appVersion: "v0.3.72"
-annotations:
- category: Productivity
- licenses: MIT
- images: |
- - name: webserver
- image: docker.io/danswer/danswer-web-server:v0.3.72
- - name: background
- image: docker.io/danswer/danswer-backend:v0.3.72
- - name: vespa
- image: vespaengine/vespa:8.277.17
-dependencies:
- - name: postgresql
- version: 14.3.1
- repository: https://charts.bitnami.com/bitnami
- condition: postgresql.enabled
- - name: vespa
- version: 0.2.3
- repository: https://unoplat.github.io/vespa-helm-charts
- condition: vespa.enabled
- - name: nginx
- version: 15.14.0
- repository: oci://registry-1.docker.io/bitnamicharts
- condition: nginx.enabled
-
-
-
\ No newline at end of file
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/_helpers.tpl b/deployment/kubernetes/charts/danswer-stack/templates/_helpers.tpl
deleted file mode 100644
index 4e6672fd677..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/_helpers.tpl
+++ /dev/null
@@ -1,62 +0,0 @@
-{{/*
-Expand the name of the chart.
-*/}}
-{{- define "danswer-stack.name" -}}
-{{- default .Chart.Name .Values.nameOverride | trunc 63 | trimSuffix "-" }}
-{{- end }}
-
-{{/*
-Create a default fully qualified app name.
-We truncate at 63 chars because some Kubernetes name fields are limited to this (by the DNS naming spec).
-If release name contains chart name it will be used as a full name.
-*/}}
-{{- define "danswer-stack.fullname" -}}
-{{- if .Values.fullnameOverride }}
-{{- .Values.fullnameOverride | trunc 63 | trimSuffix "-" }}
-{{- else }}
-{{- $name := default .Chart.Name .Values.nameOverride }}
-{{- if contains $name .Release.Name }}
-{{- .Release.Name | trunc 63 | trimSuffix "-" }}
-{{- else }}
-{{- printf "%s-%s" .Release.Name $name | trunc 63 | trimSuffix "-" }}
-{{- end }}
-{{- end }}
-{{- end }}
-
-{{/*
-Create chart name and version as used by the chart label.
-*/}}
-{{- define "danswer-stack.chart" -}}
-{{- printf "%s-%s" .Chart.Name .Chart.Version | replace "+" "_" | trunc 63 | trimSuffix "-" }}
-{{- end }}
-
-{{/*
-Common labels
-*/}}
-{{- define "danswer-stack.labels" -}}
-helm.sh/chart: {{ include "danswer-stack.chart" . }}
-{{ include "danswer-stack.selectorLabels" . }}
-{{- if .Chart.AppVersion }}
-app.kubernetes.io/version: {{ .Chart.AppVersion | quote }}
-{{- end }}
-app.kubernetes.io/managed-by: {{ .Release.Service }}
-{{- end }}
-
-{{/*
-Selector labels
-*/}}
-{{- define "danswer-stack.selectorLabels" -}}
-app.kubernetes.io/name: {{ include "danswer-stack.name" . }}
-app.kubernetes.io/instance: {{ .Release.Name }}
-{{- end }}
-
-{{/*
-Create the name of the service account to use
-*/}}
-{{- define "danswer-stack.serviceAccountName" -}}
-{{- if .Values.serviceAccount.create }}
-{{- default (include "danswer-stack.fullname" .) .Values.serviceAccount.name }}
-{{- else }}
-{{- default "default" .Values.serviceAccount.name }}
-{{- end }}
-{{- end }}
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/api-connector-pvc.yaml b/deployment/kubernetes/charts/danswer-stack/templates/api-connector-pvc.yaml
deleted file mode 100644
index 7dee7f368ff..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/api-connector-pvc.yaml
+++ /dev/null
@@ -1,19 +0,0 @@
-{{- if and .Values.persistence.api.connector.enabled}}
-apiVersion: v1
-kind: PersistentVolumeClaim
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-api-connector
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- accessModes:
- {{- range .Values.persistence.api.connector.accessModes }}
- - {{ . | quote }}
- {{- end }}
- resources:
- requests:
- storage: {{ .Values.persistence.api.connector.size | quote }}
- {{- with .Values.persistence.api.connector.storageClassName }}
- storageClassName: {{ . }}
- {{- end }}
-{{- end }}
\ No newline at end of file
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/api-deployment.yaml b/deployment/kubernetes/charts/danswer-stack/templates/api-deployment.yaml
deleted file mode 100644
index da43266baaf..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/api-deployment.yaml
+++ /dev/null
@@ -1,79 +0,0 @@
-apiVersion: apps/v1
-kind: Deployment
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-api
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- {{- if not .Values.api.autoscaling.enabled }}
- replicas: {{ .Values.api.replicaCount }}
- {{- end }}
- selector:
- matchLabels:
- {{- include "danswer-stack.selectorLabels" . | nindent 6 }}
- {{- if .Values.api.deploymentLabels }}
- {{- toYaml .Values.api.deploymentLabels | nindent 6 }}
- {{- end }}
- template:
- metadata:
- {{- with .Values.api.podAnnotations }}
- annotations:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- labels:
- {{- include "danswer-stack.labels" . | nindent 8 }}
- {{- with .Values.api.podLabels }}
- {{- toYaml . | nindent 8 }}
- {{- end }}
- spec:
- {{- with .Values.imagePullSecrets }}
- imagePullSecrets:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- serviceAccountName: {{ include "danswer-stack.serviceAccountName" . }}
- securityContext:
- {{- toYaml .Values.api.podSecurityContext | nindent 8 }}
- containers:
- - name: api-server
- securityContext:
- {{- toYaml .Values.api.securityContext | nindent 12 }}
- image: "{{ .Values.api.image.repository }}:{{ .Values.api.image.tag | default .Chart.AppVersion }}"
- imagePullPolicy: {{ .Values.api.image.pullPolicy }}
- command:
- - "/bin/sh"
- - "-c"
- - |
- alembic upgrade head &&
- echo "Starting Danswer Api Server" &&
- uvicorn danswer.main:app --host 0.0.0.0 --port 8080
- ports:
- - name: api-server-port
- containerPort: {{ .Values.api.service.port }}
- protocol: TCP
- resources:
- {{- toYaml .Values.api.resources | nindent 12 }}
- envFrom:
- - configMapRef:
- name: {{ .Values.config.envConfigMapName }}
- volumeMounts:
- - name: dynamic-storage
- mountPath: /home/storage
- - name: file-connector-storage
- mountPath: /home/file_connector_storage
- {{- if .Values.api.volumeMounts }}
- {{- .Values.api.volumeMounts | toYaml | nindent 12}}
- {{- end }}
- volumes:
- - name: dynamic-storage
- persistentVolumeClaim:
- claimName: {{ include "danswer-stack.fullname" . }}-api-dynamic
- - name: file-connector-storage
- {{- if .Values.persistence.api.connector.enabled }}
- persistentVolumeClaim:
- claimName: {{ include "danswer-stack.fullname" . }}-api-connector
- {{- else }}
- emptyDir: { }
- {{- end }}
- {{- if .Values.api.volumes }}
- {{- .Values.api.volumes | toYaml | nindent 8}}
- {{- end }}
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/api-hpa.yaml b/deployment/kubernetes/charts/danswer-stack/templates/api-hpa.yaml
deleted file mode 100644
index 378c39715ad..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/api-hpa.yaml
+++ /dev/null
@@ -1,32 +0,0 @@
-{{- if .Values.api.autoscaling.enabled }}
-apiVersion: autoscaling/v2
-kind: HorizontalPodAutoscaler
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-api
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- scaleTargetRef:
- apiVersion: apps/v1
- kind: Deployment
- name: {{ include "danswer-stack.fullname" . }}
- minReplicas: {{ .Values.api.autoscaling.minReplicas }}
- maxReplicas: {{ .Values.api.autoscaling.maxReplicas }}
- metrics:
- {{- if .Values.api.autoscaling.targetCPUUtilizationPercentage }}
- - type: Resource
- resource:
- name: cpu
- target:
- type: Utilization
- averageUtilization: {{ .Values.api.autoscaling.targetCPUUtilizationPercentage }}
- {{- end }}
- {{- if .Values.api.autoscaling.targetMemoryUtilizationPercentage }}
- - type: Resource
- resource:
- name: memory
- target:
- type: Utilization
- averageUtilization: {{ .Values.api.autoscaling.targetMemoryUtilizationPercentage }}
- {{- end }}
-{{- end }}
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/api-pvc.yaml b/deployment/kubernetes/charts/danswer-stack/templates/api-pvc.yaml
deleted file mode 100644
index 8d155d0ca45..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/api-pvc.yaml
+++ /dev/null
@@ -1,19 +0,0 @@
-{{- if and .Values.persistence.api.dynamic.enabled}}
-apiVersion: v1
-kind: PersistentVolumeClaim
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-api-dynamic
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- accessModes:
- {{- range .Values.persistence.api.dynamic.accessModes }}
- - {{ . | quote }}
- {{- end }}
- resources:
- requests:
- storage: {{ .Values.persistence.api.dynamic.size | quote }}
- {{- with .Values.persistence.api.dynamic.storageClassName }}
- storageClassName: {{ . }}
- {{- end }}
-{{- end }}
\ No newline at end of file
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/api-service.yaml b/deployment/kubernetes/charts/danswer-stack/templates/api-service.yaml
deleted file mode 100644
index 820e53d1076..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/api-service.yaml
+++ /dev/null
@@ -1,21 +0,0 @@
-apiVersion: v1
-kind: Service
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-api
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
- {{- if .Values.api.deploymentLabels }}
- {{- toYaml .Values.api.deploymentLabels | nindent 4 }}
- {{- end }}
-spec:
- type: {{ .Values.api.service.type }}
- ports:
- - port: {{ .Values.api.service.port }}
- targetPort: api-server-port
- protocol: TCP
- name: api-server-port
- selector:
- {{- include "danswer-stack.selectorLabels" . | nindent 4 }}
- {{- if .Values.api.deploymentLabels }}
- {{- toYaml .Values.api.deploymentLabels | nindent 4 }}
- {{- end }}
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/background-connector-pvc.yaml b/deployment/kubernetes/charts/danswer-stack/templates/background-connector-pvc.yaml
deleted file mode 100644
index 3bf00905769..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/background-connector-pvc.yaml
+++ /dev/null
@@ -1,19 +0,0 @@
-{{- if and .Values.persistence.api.connector.enabled}}
-apiVersion: v1
-kind: PersistentVolumeClaim
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-bg-connector
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- accessModes:
- {{- range .Values.persistence.background.connector.accessModes }}
- - {{ . | quote }}
- {{- end }}
- resources:
- requests:
- storage: {{ .Values.persistence.background.connector.size | quote }}
- {{- with .Values.persistence.background.connector.storageClassName }}
- storageClassName: {{ . }}
- {{- end }}
-{{- end }}
\ No newline at end of file
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/background-deployment.yaml b/deployment/kubernetes/charts/danswer-stack/templates/background-deployment.yaml
deleted file mode 100644
index 7e795c2a68c..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/background-deployment.yaml
+++ /dev/null
@@ -1,76 +0,0 @@
-apiVersion: apps/v1
-kind: Deployment
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-background
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- {{- if not .Values.background.autoscaling.enabled }}
- replicas: {{ .Values.background.replicaCount }}
- {{- end }}
- selector:
- matchLabels:
- {{- include "danswer-stack.selectorLabels" . | nindent 6 }}
- {{- if .Values.background.deploymentLabels }}
- {{- toYaml .Values.background.deploymentLabels | nindent 6 }}
- {{- end }}
- template:
- metadata:
- {{- with .Values.background.podAnnotations }}
- annotations:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- labels:
- {{- include "danswer-stack.labels" . | nindent 8 }}
- {{- with .Values.background.podLabels }}
- {{- toYaml . | nindent 8 }}
- {{- end }}
- spec:
- {{- with .Values.imagePullSecrets }}
- imagePullSecrets:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- serviceAccountName: {{ include "danswer-stack.serviceAccountName" . }}
- securityContext:
- {{- toYaml .Values.background.podSecurityContext | nindent 8 }}
- containers:
- - name: background
- securityContext:
- {{- toYaml .Values.background.securityContext | nindent 12 }}
- image: "{{ .Values.background.image.repository }}:{{ .Values.background.image.tag | default .Chart.AppVersion }}"
- imagePullPolicy: {{ .Values.background.image.pullPolicy }}
- command: ["/usr/bin/supervisord"]
- resources:
- {{- toYaml .Values.background.resources | nindent 12 }}
- envFrom:
- - configMapRef:
- name: {{ .Values.config.envConfigMapName }}
- env:
- - name: ENABLE_MINI_CHUNK
- value: "{{ .Values.background.enableMiniChunk }}"
- volumeMounts:
- - name: dynamic-storage
- mountPath: /home/storage
- - name: connector-storage
- mountPath: /home/file_connector_storage
- {{- if .Values.background.volumeMounts }}
- {{- .Values.background.volumeMounts | toYaml | nindent 12}}
- {{- end }}
- volumes:
- - name: dynamic-storage
- {{- if .Values.persistence.background.dynamic.enabled }}
- persistentVolumeClaim:
- claimName: {{ include "danswer-stack.fullname" . }}-bg-dynamic
- {{- else }}
- emptyDir: { }
- {{- end }}
- - name: connector-storage
- {{- if .Values.persistence.background.connector.enabled }}
- persistentVolumeClaim:
- claimName: {{ include "danswer-stack.fullname" . }}-bg-connector
- {{- else }}
- emptyDir: { }
- {{- end }}
- {{- if .Values.background.volumes }}
- {{- .Values.background.volumes | toYaml | nindent 8}}
- {{- end }}
\ No newline at end of file
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/background-dynamic-pvc.yaml b/deployment/kubernetes/charts/danswer-stack/templates/background-dynamic-pvc.yaml
deleted file mode 100644
index 8cbadd40ac0..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/background-dynamic-pvc.yaml
+++ /dev/null
@@ -1,19 +0,0 @@
-{{- if and .Values.persistence.api.dynamic.enabled}}
-apiVersion: v1
-kind: PersistentVolumeClaim
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-bg-dynamic
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- accessModes:
- {{- range .Values.persistence.background.dynamic.accessModes }}
- - {{ . | quote }}
- {{- end }}
- resources:
- requests:
- storage: {{ .Values.persistence.background.dynamic.size | quote }}
- {{- with .Values.persistence.background.dynamic.storageClassName }}
- storageClassName: {{ . }}
- {{- end }}
-{{- end }}
\ No newline at end of file
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/background-hpa.yaml b/deployment/kubernetes/charts/danswer-stack/templates/background-hpa.yaml
deleted file mode 100644
index 009daf10f05..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/background-hpa.yaml
+++ /dev/null
@@ -1,32 +0,0 @@
-{{- if .Values.background.autoscaling.enabled }}
-apiVersion: autoscaling/v2
-kind: HorizontalPodAutoscaler
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-background
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- scaleTargetRef:
- apiVersion: apps/v1
- kind: Deployment
- name: {{ include "danswer-stack.fullname" . }}
- minReplicas: {{ .Values.background.autoscaling.minReplicas }}
- maxReplicas: {{ .Values.background.autoscaling.maxReplicas }}
- metrics:
- {{- if .Values.background.autoscaling.targetCPUUtilizationPercentage }}
- - type: Resource
- resource:
- name: cpu
- target:
- type: Utilization
- averageUtilization: {{ .Values.background.autoscaling.targetCPUUtilizationPercentage }}
- {{- end }}
- {{- if .Values.background.autoscaling.targetMemoryUtilizationPercentage }}
- - type: Resource
- resource:
- name: memory
- target:
- type: Utilization
- averageUtilization: {{ .Values.background.autoscaling.targetMemoryUtilizationPercentage }}
- {{- end }}
-{{- end }}
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/configmap.yaml b/deployment/kubernetes/charts/danswer-stack/templates/configmap.yaml
deleted file mode 100755
index b9812342f01..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/configmap.yaml
+++ /dev/null
@@ -1,15 +0,0 @@
-apiVersion: v1
-kind: ConfigMap
-metadata:
- name: env-configmap
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-data:
- INTERNAL_URL: "{{ include "danswer-stack.fullname" . }}-api:80"
- POSTGRES_HOST: {{ .Release.Name }}-postgresql
- VESPA_HOST: "document-index-service"
- MODEL_SERVER_HOST: "{{ include "danswer-stack.fullname" . }}-inference-model-service"
- INDEXING_MODEL_SERVER_HOST: "{{ include "danswer-stack.fullname" . }}-indexing-model-service"
-{{- range $key, $value := .Values.configMap }}
- {{ $key }}: "{{ $value }}"
-{{- end }}
\ No newline at end of file
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/danswer-secret.yaml b/deployment/kubernetes/charts/danswer-stack/templates/danswer-secret.yaml
deleted file mode 100644
index 2586048f05b..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/danswer-secret.yaml
+++ /dev/null
@@ -1,11 +0,0 @@
-apiVersion: v1
-kind: Secret
-metadata:
- name: danswer-secrets
-type: Opaque
-data:
- postgres_user: cG9zdGdyZXM= # "postgres" base64 encoded
- postgres_password: cGFzc3dvcmQ= # "password" base64 encoded
- postgres-password: cGFzc3dvcmQ=
- google_oauth_client_id: # You will need to provide this, use echo -n "your-client-id" | base64
- google_oauth_client_secret: # You
\ No newline at end of file
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/serviceaccount.yaml b/deployment/kubernetes/charts/danswer-stack/templates/serviceaccount.yaml
deleted file mode 100644
index afd351217ba..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/serviceaccount.yaml
+++ /dev/null
@@ -1,13 +0,0 @@
-{{- if .Values.serviceAccount.create -}}
-apiVersion: v1
-kind: ServiceAccount
-metadata:
- name: {{ include "danswer-stack.serviceAccountName" . }}
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
- {{- with .Values.serviceAccount.annotations }}
- annotations:
- {{- toYaml . | nindent 4 }}
- {{- end }}
-automountServiceAccountToken: {{ .Values.serviceAccount.automount }}
-{{- end }}
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/tests/test-connection.yaml b/deployment/kubernetes/charts/danswer-stack/templates/tests/test-connection.yaml
deleted file mode 100644
index 60fbd1054c1..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/tests/test-connection.yaml
+++ /dev/null
@@ -1,15 +0,0 @@
-apiVersion: v1
-kind: Pod
-metadata:
- name: "{{ include "danswer-stack.fullname" . }}-test-connection"
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
- annotations:
- "helm.sh/hook": test
-spec:
- containers:
- - name: wget
- image: busybox
- command: ['wget']
- args: ['{{ include "danswer-stack.fullname" . }}:{{ .Values.webserver.service.port }}']
- restartPolicy: Never
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/webserver-deployment.yaml b/deployment/kubernetes/charts/danswer-stack/templates/webserver-deployment.yaml
deleted file mode 100644
index 20225a3ccd7..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/webserver-deployment.yaml
+++ /dev/null
@@ -1,58 +0,0 @@
-apiVersion: apps/v1
-kind: Deployment
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-webserver
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- {{- if not .Values.webserver.autoscaling.enabled }}
- replicas: {{ .Values.webserver.replicaCount }}
- {{- end }}
- selector:
- matchLabels:
- {{- include "danswer-stack.selectorLabels" . | nindent 6 }}
- {{- if .Values.webserver.deploymentLabels }}
- {{- toYaml .Values.webserver.deploymentLabels | nindent 6 }}
- {{- end }}
- template:
- metadata:
- {{- with .Values.webserver.podAnnotations }}
- annotations:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- labels:
- {{- include "danswer-stack.labels" . | nindent 8 }}
- {{- with .Values.webserver.podLabels }}
- {{- toYaml . | nindent 8 }}
- {{- end }}
- spec:
- {{- with .Values.imagePullSecrets }}
- imagePullSecrets:
- {{- toYaml . | nindent 8 }}
- {{- end }}
- serviceAccountName: {{ include "danswer-stack.serviceAccountName" . }}
- securityContext:
- {{- toYaml .Values.webserver.podSecurityContext | nindent 8 }}
- containers:
- - name: web-server
- securityContext:
- {{- toYaml .Values.webserver.securityContext | nindent 12 }}
- image: "{{ .Values.webserver.image.repository }}:{{ .Values.webserver.image.tag | default .Chart.AppVersion }}"
- imagePullPolicy: {{ .Values.webserver.image.pullPolicy }}
- ports:
- - name: http
- containerPort: {{ .Values.webserver.service.port }}
- protocol: TCP
- resources:
- {{- toYaml .Values.webserver.resources | nindent 12 }}
- envFrom:
- - configMapRef:
- name: {{ .Values.config.envConfigMapName }}
- {{- with .Values.webserver.volumeMounts }}
- volumeMounts:
- {{- toYaml . | nindent 12 }}
- {{- end }}
- {{- with .Values.webserver.volumes }}
- volumes:
- {{- toYaml . | nindent 8 }}
- {{- end }}
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/webserver-hpa.yaml b/deployment/kubernetes/charts/danswer-stack/templates/webserver-hpa.yaml
deleted file mode 100644
index b46820a7fac..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/webserver-hpa.yaml
+++ /dev/null
@@ -1,32 +0,0 @@
-{{- if .Values.webserver.autoscaling.enabled }}
-apiVersion: autoscaling/v2
-kind: HorizontalPodAutoscaler
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-webserver
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
-spec:
- scaleTargetRef:
- apiVersion: apps/v1
- kind: Deployment
- name: {{ include "danswer-stack.fullname" . }}
- minReplicas: {{ .Values.webserver.autoscaling.minReplicas }}
- maxReplicas: {{ .Values.webserver.autoscaling.maxReplicas }}
- metrics:
- {{- if .Values.webserver.autoscaling.targetCPUUtilizationPercentage }}
- - type: Resource
- resource:
- name: cpu
- target:
- type: Utilization
- averageUtilization: {{ .Values.webserver.autoscaling.targetCPUUtilizationPercentage }}
- {{- end }}
- {{- if .Values.webserver.autoscaling.targetMemoryUtilizationPercentage }}
- - type: Resource
- resource:
- name: memory
- target:
- type: Utilization
- averageUtilization: {{ .Values.webserver.autoscaling.targetMemoryUtilizationPercentage }}
- {{- end }}
-{{- end }}
diff --git a/deployment/kubernetes/charts/danswer-stack/templates/webserver-service.yaml b/deployment/kubernetes/charts/danswer-stack/templates/webserver-service.yaml
deleted file mode 100644
index 3e33566fce1..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/templates/webserver-service.yaml
+++ /dev/null
@@ -1,21 +0,0 @@
-apiVersion: v1
-kind: Service
-metadata:
- name: {{ include "danswer-stack.fullname" . }}-webserver
- labels:
- {{- include "danswer-stack.labels" . | nindent 4 }}
- {{- if .Values.webserver.deploymentLabels }}
- {{- toYaml .Values.webserver.deploymentLabels | nindent 4 }}
- {{- end }}
-spec:
- type: {{ .Values.webserver.service.type }}
- ports:
- - port: {{ .Values.webserver.service.port }}
- targetPort: http
- protocol: TCP
- name: http
- selector:
- {{- include "danswer-stack.selectorLabels" . | nindent 4 }}
- {{- if .Values.webserver.deploymentLabels }}
- {{- toYaml .Values.webserver.deploymentLabels | nindent 4 }}
- {{- end }}
diff --git a/deployment/kubernetes/charts/danswer-stack/values.yaml b/deployment/kubernetes/charts/danswer-stack/values.yaml
deleted file mode 100644
index 7a825d62367..00000000000
--- a/deployment/kubernetes/charts/danswer-stack/values.yaml
+++ /dev/null
@@ -1,463 +0,0 @@
-# Default values for danswer-stack.
-# This is a YAML-formatted file.
-# Declare variables to be passed into your templates.
-
-imagePullSecrets: []
-nameOverride: ""
-fullnameOverride: ""
-
-inferenceCapability:
- service:
- name: inference-model-server-service
- type: ClusterIP
- port: 9000
- pvc:
- name: inference-model-pvc
- accessModes:
- - ReadWriteOnce
- storage: 3Gi
- deployment:
- name: inference-model-server-deployment
- replicas: 1
- labels:
- - key: app
- value: inference-model-server
- image:
- repository: danswer/danswer-model-server
- tag: latest
- pullPolicy: IfNotPresent
- command: ["uvicorn", "model_server.main:app", "--host", "0.0.0.0", "--port", "9000"]
- port: 9000
- volumeMounts:
- - name: inference-model-storage
- mountPath: /root/.cache
- volumes:
- - name: inference-model-storage
- persistentVolumeClaim:
- claimName: inference-model-pvc
- podLabels:
- - key: app
- value: inference-model-server
-
-
-
-config:
- envConfigMapName: env-configmap
-
-serviceAccount:
- # Specifies whether a service account should be created
- create: false
- # Automatically mount a ServiceAccount's API credentials?
- automount: true
- # Annotations to add to the service account
- annotations: {}
- # The name of the service account to use.
- # If not set and create is true, a name is generated using the fullname template
- name: ""
-
-indexCapability:
- service:
- type: ClusterIP
- port: 9000
- name: indexing-model-server-port
- deploymentLabels:
- app: indexing-model-server
- podLabels:
- app: indexing-model-server
- indexingOnly: "True"
- podAnnotations: {}
- volumeMounts:
- - name: indexing-model-storage
- mountPath: /root/.cache
- volumes:
- - name: indexing-model-storage
- persistentVolumeClaim:
- claimName: indexing-model-storage
- indexingModelPVC:
- name: indexing-model-storage
- accessMode: "ReadWriteOnce"
- storage: "3Gi"
-
-postgresql:
- primary:
- persistence:
- size: 1Gi
- enabled: true
- auth:
- existingSecret: danswer-secrets
- # secretKeys:
- # adminPasswordKey: postgres_password
-
-nginx:
- containerPorts:
- http: 1024
- extraEnvVars:
- - name: DOMAIN
- value: localhost
- service:
- ports:
- http: 80
- danswer: 3000
- targetPort:
- http: http
- danswer: http
-
- existingServerBlockConfigmap: danswer-nginx-conf
-
-webserver:
- replicaCount: 1
- image:
- repository: danswer/danswer-web-server
- pullPolicy: IfNotPresent
- # Overrides the image tag whose default is the chart appVersion.
- tag: ""
- deploymentLabels:
- app: web-server
- podAnnotations: {}
- podLabels:
- app: web-server
- podSecurityContext: {}
- # fsGroup: 2000
-
- securityContext: {}
- # capabilities:
- # drop:
- # - ALL
- # readOnlyRootFilesystem: true
- # runAsNonRoot: true
- # runAsUser: 1000
-
- service:
- type: ClusterIP
- port: 3000
-
- resources: {}
- # We usually recommend not to specify default resources and to leave this as a conscious
- # choice for the user. This also increases chances charts run on environments with little
- # resources, such as Minikube. If you do want to specify resources, uncomment the following
- # lines, adjust them as necessary, and remove the curly braces after 'resources:'.
- # limits:
- # cpu: 100m
- # memory: 128Mi
- # requests:
- # cpu: 100m
- # memory: 128Mi
-
- autoscaling:
- enabled: false
- minReplicas: 1
- maxReplicas: 100
- targetCPUUtilizationPercentage: 80
- # targetMemoryUtilizationPercentage: 80
-
- # Additional volumes on the output Deployment definition.
- volumes: []
- # - name: foo
- # secret:
- # secretName: mysecret
- # optional: false
-
- # Additional volumeMounts on the output Deployment definition.
- volumeMounts: []
- # - name: foo
- # mountPath: "/etc/foo"
- # readOnly: true
-
- nodeSelector: {}
- tolerations: []
- affinity: {}
-
-api:
- replicaCount: 1
- image:
- repository: danswer/danswer-backend
- pullPolicy: IfNotPresent
- # Overrides the image tag whose default is the chart appVersion.
- tag: ""
- deploymentLabels:
- app: api-server
- podAnnotations: {}
- podLabels:
- scope: danswer-backend
- app: api-server
-
- podSecurityContext: {}
- # fsGroup: 2000
-
- securityContext: {}
- # capabilities:
- # drop:
- # - ALL
- # readOnlyRootFilesystem: true
- # runAsNonRoot: true
- # runAsUser: 1000
-
- service:
- type: ClusterIP
- port: 8080
-
- resources:
- # We usually recommend not to specify default resources and to leave this as a conscious
- # choice for the user. This also increases chances charts run on environments with little
- # resources, such as Minikube. If you do want to specify resources, uncomment the following
- # lines, adjust them as necessary, and remove the curly braces after 'resources:'.
- requests:
- cpu: 500m
- memory: 128Mi
- limits:
- cpu: 900m
- memory: 512Mi
-
- autoscaling:
- enabled: false
- minReplicas: 1
- maxReplicas: 100
- targetCPUUtilizationPercentage: 80
- # targetMemoryUtilizationPercentage: 80
-
- # Additional volumes on the output Deployment definition.
- volumes: []
- # - name: foo
- # secret:
- # secretName: mysecret
- # optional: false
-
- # Additional volumeMounts on the output Deployment definition.
- volumeMounts: []
- # - name: foo
- # mountPath: "/etc/foo"
- # readOnly: true
-
- nodeSelector: {}
- tolerations: []
-
-
-background:
- replicaCount: 1
- image:
- repository: danswer/danswer-backend
- pullPolicy: IfNotPresent
- # Overrides the image tag whose default is the chart appVersion.
- tag: latest
- podAnnotations: {}
- podLabels:
- scope: danswer-backend
- app: background
- deploymentLabels:
- app: background
- podSecurityContext: {}
- # fsGroup: 2000
-
- securityContext: {}
- # capabilities:
- # drop:
- # - ALL
- # readOnlyRootFilesystem: true
- # runAsNonRoot: true
- # runAsUser: 1000
- enableMiniChunk: "true"
- resources:
- # We usually recommend not to specify default resources and to leave this as a conscious
- # choice for the user. This also increases chances charts run on environments with little
- # resources, such as Minikube. If you do want to specify resources, uncomment the following
- # lines, adjust them as necessary, and remove the curly braces after 'resources:'.
- requests:
- cpu: 500m
- memory: 500Mi
- limits:
- cpu: 900m
- memory: 800Mi
-
- autoscaling:
- enabled: false
- minReplicas: 1
- maxReplicas: 100
- targetCPUUtilizationPercentage: 80
- # targetMemoryUtilizationPercentage: 80
-
- # Additional volumes on the output Deployment definition.
- volumes: []
- # - name: foo
- # secret:
- # secretName: mysecret
- # optional: false
-
- # Additional volumeMounts on the output Deployment definition.
- volumeMounts: []
- # - name: foo
- # mountPath: "/etc/foo"
- # readOnly: true
-
- nodeSelector: {}
- tolerations: []
-
-vespa:
- replicaCount: 1
- image:
- repository: vespa
- pullPolicy: IfNotPresent
- tag: "8.277.17"
- podAnnotations: {}
- podLabels:
- app: vespa
- app.kubernetes.io/instance: danswer-stack-kn
- app.kubernetes.io/name: vespa
- enabled: true
-
- podSecurityContext: {}
- # fsGroup: 2000
-
- securityContext:
- privileged: true
- runAsUser: 0
- # capabilities:
- # drop:
- # - ALL
- # readOnlyRootFilesystem: true
- # runAsNonRoot: true
- # runAsUser: 1000
-
- resources:
- # We usually recommend not to specify default resources and to leave this as a conscious
- # choice for the user. This also increases chances charts run on environments with little
- # resources, such as Minikube. If you do want to specify resources, uncomment the following
- # lines, adjust them as necessary, and remove the curly braces after 'resources:'.
- # requests:
- # cpu: 1500m
- # memory: 4000Mi
- # # limits:
- # # cpu: 100m
- # # memory: 128Mi
-
- nodeSelector: {}
- tolerations: []
- affinity: {}
-
-
-#ingress:
-# enabled: false
-# className: ""
-# annotations: {}
-# # kubernetes.io/ingress.class: nginx
-# # kubernetes.io/tls-acme: "true"
-# hosts:
-# - host: chart-example.local
-# paths:
-# - path: /
-# pathType: ImplementationSpecific
-# tls: []
-# # - secretName: chart-example-tls
-# # hosts:
-# # - chart-example.local
-
-persistence:
- vespa:
- enabled: true
- existingClaim: ""
- storageClassName: ""
- accessModes:
- - ReadWriteOnce
- size: 1Gi
- connector:
- enabled: true
- existingClaim: ""
- storageClassName: ""
- accessModes:
- - ReadWriteOnce
- size: 1Gi
- api:
- connector:
- enabled: true
- storageClassName: ""
- accessModes:
- - ReadWriteOnce
- size: 1Gi
- dynamic:
- enabled: true
- storageClassName: ""
- accessModes:
- - ReadWriteOnce
- size: 1Gi
- background:
- connector:
- enabled: true
- storageClassName: ""
- accessModes:
- - ReadWriteOnce
- size: 1Gi
- dynamic:
- enabled: true
- storageClassName: ""
- accessModes:
- - ReadWriteOnce
- size: 1Gi
-
-configMap:
- AUTH_TYPE: "disabled" # Change this for production uses unless Danswer is only accessible behind VPN
- SESSION_EXPIRE_TIME_SECONDS: "86400" # 1 Day Default
- VALID_EMAIL_DOMAINS: "" # Can be something like danswer.ai, as an extra double-check
- SMTP_SERVER: "" # For sending verification emails, if unspecified then defaults to 'smtp.gmail.com'
- SMTP_PORT: "" # For sending verification emails, if unspecified then defaults to '587'
- SMTP_USER: "" # 'your-email@company.com'
- SMTP_PASS: "" # 'your-gmail-password'
- EMAIL_FROM: "" # 'your-email@company.com' SMTP_USER missing used instead
- # Gen AI Settings
- GEN_AI_MODEL_PROVIDER: ""
- GEN_AI_MODEL_VERSION: ""
- FAST_GEN_AI_MODEL_VERSION: ""
- GEN_AI_API_KEY: ""
- GEN_AI_API_ENDPOINT: ""
- GEN_AI_API_VERSION: ""
- GEN_AI_LLM_PROVIDER_TYPE: ""
- GEN_AI_MAX_TOKENS: ""
- QA_TIMEOUT: "60"
- MAX_CHUNKS_FED_TO_CHAT: ""
- DISABLE_LLM_FILTER_EXTRACTION: ""
- DISABLE_LLM_CHUNK_FILTER: ""
- DISABLE_LLM_CHOOSE_SEARCH: ""
- DISABLE_LLM_QUERY_REPHRASE: ""
- # Query Options
- DOC_TIME_DECAY: ""
- HYBRID_ALPHA: ""
- EDIT_KEYWORD_QUERY: ""
- MULTILINGUAL_QUERY_EXPANSION: ""
- QA_PROMPT_OVERRIDE: ""
- # Don't change the NLP models unless you know what you're doing
- DOCUMENT_ENCODER_MODEL: ""
- NORMALIZE_EMBEDDINGS: ""
- ASYM_QUERY_PREFIX: ""
- ASYM_PASSAGE_PREFIX: ""
- ENABLE_RERANKING_REAL_TIME_FLOW: ""
- ENABLE_RERANKING_ASYNC_FLOW: ""
-
- MODEL_SERVER_PORT: ""
-
- MIN_THREADS_ML_MODELS: ""
- # Indexing Configs
- NUM_INDEXING_WORKERS: ""
- DISABLE_INDEX_UPDATE_ON_SWAP: ""
- DASK_JOB_CLIENT_ENABLED: ""
- CONTINUE_ON_CONNECTOR_FAILURE: ""
- EXPERIMENTAL_CHECKPOINTING_ENABLED: ""
- CONFLUENCE_CONNECTOR_LABELS_TO_SKIP: ""
- JIRA_API_VERSION: ""
- GONG_CONNECTOR_START_TIME: ""
- NOTION_CONNECTOR_ENABLE_RECURSIVE_PAGE_LOOKUP: ""
- # DanswerBot SlackBot Configs
- DANSWER_BOT_SLACK_APP_TOKEN: ""
- DANSWER_BOT_SLACK_BOT_TOKEN: ""
- DANSWER_BOT_DISABLE_DOCS_ONLY_ANSWER: ""
- DANSWER_BOT_DISPLAY_ERROR_MSGS: ""
- DANSWER_BOT_RESPOND_EVERY_CHANNEL: ""
- DANSWER_BOT_DISABLE_COT: "" # Currently unused
- NOTIFY_SLACKBOT_NO_ANSWER: ""
- # Logging
- # Optional Telemetry, please keep it on (nothing sensitive is collected)? <3
- # https://docs.danswer.dev/more/telemetry
- DISABLE_TELEMETRY: ""
- LOG_LEVEL: ""
- LOG_ALL_MODEL_INTERACTIONS: ""
- LOG_VESPA_TIMING_INFORMATION: ""
- # Shared or Non-backend Related
- WEB_DOMAIN: "http://localhost:3000" # for web server and api server
- DOMAIN: "localhost" # for nginx
diff --git a/deployment/kubernetes/persistent-volumes.yaml b/deployment/kubernetes/persistent-volumes.yaml
deleted file mode 100644
index 8376b98e697..00000000000
--- a/deployment/kubernetes/persistent-volumes.yaml
+++ /dev/null
@@ -1,21 +0,0 @@
-apiVersion: v1
-kind: PersistentVolumeClaim
-metadata:
- name: dynamic-pvc
-spec:
- accessModes:
- - ReadWriteOnce
- resources:
- requests:
- storage: 1Gi
----
-apiVersion: v1
-kind: PersistentVolumeClaim
-metadata:
- name: file-connector-pvc
-spec:
- accessModes:
- - ReadWriteOnce
- resources:
- requests:
- storage: 1Gi
diff --git a/deployment/kubernetes/postgres-service-deployment.yaml b/deployment/kubernetes/postgres-service-deployment.yaml
index f33efa2bafd..17330204c1e 100644
--- a/deployment/kubernetes/postgres-service-deployment.yaml
+++ b/deployment/kubernetes/postgres-service-deployment.yaml
@@ -54,4 +54,4 @@ spec:
resources:
requests:
# Adjust the storage request size as needed.
- storage: 1Gi
+ storage: 5Gi
diff --git a/deployment/kubernetes/secrets.yaml b/deployment/kubernetes/secrets.yaml
index 8193092f4a9..c135a29f676 100644
--- a/deployment/kubernetes/secrets.yaml
+++ b/deployment/kubernetes/secrets.yaml
@@ -7,5 +7,5 @@ type: Opaque
data:
postgres_user: cG9zdGdyZXM= # "postgres" base64 encoded
postgres_password: cGFzc3dvcmQ= # "password" base64 encoded
- google_oauth_client_id: # You will need to provide this, use echo -n "your-client-id" | base64
- google_oauth_client_secret: # You will need to provide this, use echo -n "your-client-id" | base64
+ google_oauth_client_id: ZXhhbXBsZS1jbGllbnQtaWQ= # "example-client-id" base64 encoded. You will need to provide this, use echo -n "your-client-id" | base64
+ google_oauth_client_secret: example_google_oauth_secret # "example-client-secret" base64 encoded. You will need to provide this, use echo -n "your-client-id" | base64
diff --git a/deployment/kubernetes/vespa-service-deployment.yaml b/deployment/kubernetes/vespa-service-deployment.yaml
index 4fa5aa9fac5..5016258b757 100644
--- a/deployment/kubernetes/vespa-service-deployment.yaml
+++ b/deployment/kubernetes/vespa-service-deployment.yaml
@@ -60,4 +60,4 @@ spec:
resources:
requests:
# Adjust the storage request size as needed.
- storage: 1Gi
+ storage: 5Gi
diff --git a/web/Dockerfile b/web/Dockerfile
index a0b16c75d3b..3d27813c7ae 100644
--- a/web/Dockerfile
+++ b/web/Dockerfile
@@ -54,6 +54,9 @@ ENV NEXT_PUBLIC_POSITIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_POSITIVE_PRED
ARG NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS
ENV NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS}
+ARG NEXT_PUBLIC_DISABLE_LOGOUT
+ENV NEXT_PUBLIC_DISABLE_LOGOUT=${NEXT_PUBLIC_DISABLE_LOGOUT}
+
RUN npm run build
# Step 3. Production image, copy all the files and run next
@@ -99,6 +102,9 @@ ENV NEXT_PUBLIC_POSITIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_POSITIVE_PRED
ARG NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS
ENV NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS=${NEXT_PUBLIC_NEGATIVE_PREDEFINED_FEEDBACK_OPTIONS}
+ARG NEXT_PUBLIC_DISABLE_LOGOUT
+ENV NEXT_PUBLIC_DISABLE_LOGOUT=${NEXT_PUBLIC_DISABLE_LOGOUT}
+
# Note: Don't expose ports here, Compose will handle that for us if necessary.
# If you want to run this without compose, specify the ports to
# expose via cli
diff --git a/web/public/Dropbox.png b/web/public/Dropbox.png
new file mode 100644
index 00000000000..cd83e09eb66
Binary files /dev/null and b/web/public/Dropbox.png differ
diff --git a/web/src/app/admin/assistants/AssistantEditor.tsx b/web/src/app/admin/assistants/AssistantEditor.tsx
index 8c1f01973b2..d6427a67929 100644
--- a/web/src/app/admin/assistants/AssistantEditor.tsx
+++ b/web/src/app/admin/assistants/AssistantEditor.tsx
@@ -34,6 +34,9 @@ import { DocumentSetSelectable } from "@/components/documentSet/DocumentSetSelec
import { FullLLMProvider } from "../models/llm/interfaces";
import { Option } from "@/components/Dropdown";
import { ToolSnapshot } from "@/lib/tools/interfaces";
+import { checkUserIsNoAuthUser } from "@/lib/user";
+import { addAssistantToList } from "@/lib/assistants/updateAssistantPreferences";
+import { checkLLMSupportsImageInput } from "@/lib/llm/utils";
function findSearchTool(tools: ToolSnapshot[]) {
return tools.find((tool) => tool.in_code_tool_id === "SearchTool");
@@ -43,12 +46,6 @@ function findImageGenerationTool(tools: ToolSnapshot[]) {
return tools.find((tool) => tool.in_code_tool_id === "ImageGenerationTool");
}
-function checkLLMSupportsImageGeneration(provider: string, model: string) {
- console.log(provider);
- console.log(model);
- return provider === "openai" && model === "gpt-4-turbo";
-}
-
function Label({ children }: { children: string | JSX.Element }) {
return (
Existing API Token:
++ {dropboxCredential.credential_json?.dropbox_access_token} +
+ ++ Press connect below to start the connection to your Dropbox + instance. +
+