feat(mapping-optimizer): Support in operator for mapping optimizer #5685
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This was a TODO item. But on the spans dataset, one easy to encounter situation
is a condition like `sentry_tags[key] IN (value1, value2)`. This results in a
sql like
`in((arrayElement(sentry_tags.value, indexOf(sentry_tags.key, 'key')) AS `_snuba_sentry_tags[key]`), ['value1', 'value1'])`
which scans the entire `sentry_tags.key` and `sentry_tags.value` columns. The
optimization here is to use the tags hash map which gives us a condition like
`hasAny(_sentry_tags_hash_map, array(cityHash64('environment=prod'), cityHash64('environment=production')))`.
Codecov ReportAll modified and coverable lines are covered by tests ✅
✅ All tests successful. No failed tests found Additional details and impacted files@@ Coverage Diff @@
## master #5685 +/- ##
=======================================
Coverage 89.92% 89.93%
=======================================
Files 898 898
Lines 43453 43474 +21
Branches 299 299
=======================================
+ Hits 39077 39098 +21
Misses 4334 4334
Partials 42 42 ☔ View full report in Codecov by Sentry. |
evanh
approved these changes
Mar 25, 2024
Zylphrex
commented
Mar 25, 2024
| @@ -265,7 +351,7 @@ def _get_condition_without_redundant_checks( | |||
| if tag_exist_match: | |||
| matched_tag_exists_conditions[condition_id] = tag_exist_match | |||
| if not tag_exist_match: | |||
| eq_match = self.__optimizable_pattern.match(cond) | |||
| eq_match = self.__equals_condition_pattern.match(cond) | |||
There was a problem hiding this comment.
@volokluev would you happen to know if I need to implement this removing of redundant checks for IN conditions?
There was a problem hiding this comment.
you could but it's not strictly necessary. I'm not sure how often we get those cases with IN conditions. Definitely something that can be added later
There was a problem hiding this comment.
I think this is less common on the older datasets but more likely to happen with the spans dataset as we have sentry_tags which contains some more commonly used columns.
The example I ran into was with environment. For a 24h period, it read >48GiB of data, and after applying this optimization, I saw it was reduced to <24GiB of data. On 7 day periods, the query was already timing out. So this optimization should already be helpful
There was a problem hiding this comment.
oh yes absolutely. Merge the PR. I was saying that the redundant clause optimization is probably not going to be as applicable for IN clauses
Zylphrex
deleted the
txiao/feat/support-in-operator-for-mapping-optimizers
branch
|
PR reverted: 8c6329d |
Zylphrex
added a commit
that referenced
this pull request
Mar 26, 2024
Re-apply #5685 This was a TODO item. But on the spans dataset, one easy to encounter situation is a condition like ``` sentry_tags[key] IN (value1, value2) ``` This results in a sql like ``` in((arrayElement(sentry_tags.value, indexOf(sentry_tags.key, 'key')) AS `_snuba_sentry_tags[key]`), ['value1', 'value1']) ``` which scans the entire `sentry_tags.key` and `sentry_tags.value` columns. The optimization here is to use the tags hash map which gives us a condition like ``` hasAny(_sentry_tags_hash_map, array(cityHash64('key=value1'), cityHash64('key=value1'))) ```
Zylphrex
added a commit
that referenced
this pull request
Mar 26, 2024
…5691) Re-apply #5685 This was a TODO item. But on the spans dataset, one easy to encounter situation is a condition like ``` sentry_tags[key] IN (value1, value2) ``` This results in a sql like ``` in((arrayElement(sentry_tags.value, indexOf(sentry_tags.key, 'key')) AS `_snuba_sentry_tags[key]`), ['value1', 'value1']) ``` which scans the entire `sentry_tags.key` and `sentry_tags.value` columns. The optimization here is to use the tags hash map which gives us a condition like ``` hasAny(_sentry_tags_hash_map, array(cityHash64('key=value1'), cityHash64('key=value1'))) ```
Zylphrex
added a commit
that referenced
this pull request
Mar 26, 2024
…5691) Re-apply #5685 This was a TODO item. But on the spans dataset, one easy to encounter situation is a condition like ``` sentry_tags[key] IN (value1, value2) ``` This results in a sql like ``` in((arrayElement(sentry_tags.value, indexOf(sentry_tags.key, 'key')) AS `_snuba_sentry_tags[key]`), ['value1', 'value1']) ``` which scans the entire `sentry_tags.key` and `sentry_tags.value` columns. The optimization here is to use the tags hash map which gives us a condition like ``` hasAny(_sentry_tags_hash_map, array(cityHash64('key=value1'), cityHash64('key=value1'))) ```
Zylphrex
added a commit
that referenced
this pull request
Mar 27, 2024
…5691) (#5692) Re-apply #5685 This was a TODO item. But on the spans dataset, one easy to encounter situation is a condition like ``` sentry_tags[key] IN (value1, value2) ``` This results in a sql like ``` in((arrayElement(sentry_tags.value, indexOf(sentry_tags.key, 'key')) AS `_snuba_sentry_tags[key]`), ['value1', 'value1']) ``` which scans the entire `sentry_tags.key` and `sentry_tags.value` columns. The optimization here is to use the tags hash map which gives us a condition like ``` hasAny(_sentry_tags_hash_map, array(cityHash64('key=value1'), cityHash64('key=value1'))) ```
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This was a TODO item. But on the spans dataset, one easy to encounter situation is a condition like
This results in a sql like
which scans the entire
sentry_tags.keyandsentry_tags.valuecolumns. The optimization here is to use the tags hash map which gives us a condition like