Releases: telefonicaid/fiware-cygnus
Releases · telefonicaid/fiware-cygnus
0.10.0
- [HARDENING] Cygnus user creation added to the from the sources installation guide (#558)
- [FEATURE] Add a binary implementation of the HDFS backend (#537)
- [HARDENING] Add docker as an option in the README for installing Cygnus (#552)
- [FEATURE] Add support to batch processing within a transaction in OrionHDFSSink (#554)
- [FEATURE] Add support to batch processing within a transaction in OrionMySQLSink (#568)
- [BUG] Add support to CKAN 2.4 API changes (#546)
- [BUG] Remove final line feed in the HDFS aggregation (#575)
- [FEATURE] Add some interesting fields to HDFS persisted data (servicePath in row-like modes, servicePath, entityId and entityType in column-like modes), including migration script (#544)
- [HARDENING] Document how to choose the hadoop-core dependency for compilation purposes (#559)
- [HARDENING] Add a batching timeout and accumulate the transaction IDs for right tracing (#562)
- [BUG] Ask for CKAN resources when populating a package cache after calling organization_show (#589)
- [FEATURE] Add some interesting fields to MySQL persisted data (servicePath in row-like modes, servicePath, entityId and entityType in column-like modes), including migration script (#501)
- [BUG] Fix the CKAN resource name building, removing references to old default destination (#592)
- [FEATURE] Allow enabling/disabling Hive in OrionHDFSSink (#555)
- [HARDENING] Explain the batching mechanism in the performance document (#564)
- [HARDENING] Fix the documentation regarding the installation from sources (#596)
- [BUG] Add extra fields (servicePath in row-like modes, servicePath, entityId and entityType in column-like modes) to Hive tables (#598)
- [FEATURE] Column-like persistence in OrionMongoSink (#548)
0.9.0
Changelog:
- [FEATURE] Add support for multiple file formats in HDFS; JSON is maintained and CSV is added (#303)
- [HARDENING] Add detailed information about reporting issues and contact information (#478)
- [HARDENING] Update the Quick Start Guide in order is is aligned with latest features (#490)
- [HARDENING] Remove patch supporting old versions of the CKAN API (<= 2.0) (#188)
- [FEATURE] The notified user agent is not checked anymore (#493)
- [FEATURE] Add OrionKafkaSink (#456)
- [HARDENING] Add missing configuration and documentation regarding OrionMongoSink and OrionSTHSink (#505)
- [FEATURE] Enable/disable grouping rules usage per sink (#447)
- [FEATURE] Add HiveServer2 support to OrionHDFSSink (#513)
- [FEATURE] Add Python-based clients both for HiveServer1 and HiveServer2 (#519)
- [FEATURE] Add Java-based clients both for HiveServer1 and HiveServer2 (#518)
- [HARDENING] Replace logs containing "HttpFS response" by "Server response" (#517)
- [BUG] Use a testing version of Zookeeper in OrionKafkaSink (#514)
- [BUG] Always create the Hive tables if not existing yet (#401)
- [HARDENING] Add {headers,body} structure to messages sent to Kafka (#512)
- [FEATURE] Allow selecting the table type (by service, servicePath or destination) (#540)
- [BUG] A second CKAN resource may be added to an existeng dataset in row mode (#280)
- [HARDENING] Add references to FIWARE and Cosmos Big Data Analysis GE in the README (#534)
- [BUG] entityId and entityType are now added to the CKAN resources when using OrionCKANSink in row mode (#539)
0.8.2
- [BUG] Fix many errors in the docuemtnation (#435)
- [BUG] Fix OrionMySQLLink in the README (#442)
- [BUG] Fix MySQL connections, adding a permanent connection to each database/fiware-service (#445)
- [BUG] Fix mongo-sink appearance in sth-sink configuration (both template and README) (#449)
- [BUG] Fix OrionMySQLSink, time zone is not added to the timestamp fields (#441)
- [HARDENING] Link the Quick Start Guide from the README (#454)
- [BUG] Fix the grouping rules validator, now empty fields and non-numeric ids are not allowed (#460)
- [HARDENING] Add detailed explanation about the syntax of the grouping rules (#459)
- [FEATURE] OAuth2 support for OrionHDFSSink (#483)
0.8.1
- [HARDENING] OrionHDFSSink parameters are now called "hdfs__" instead of "cosmos__" (#374)
- [HARDENING] Add a permament connection to the MySQL Backend (#364)
- [BUG] Add the original (and removed by OrionRESTHandler) slash character to fiware-servicePath (#403)
- [FEATURE] Use Json-like rules definition for pattern-based grouping (#387)
- [BUG] Set the Fiware-Service and Fiware-ServicePath maximum lenght to 50 characters (previously, it was 32 characters) (#406)
- [FEATURE] Add implementation details to the specific sinks documentation (#397)
- [BUG] Fix the origin date in OrionSTHSink (#413)
- [HARDENING] Remove unnecessary hfds_api parameter in OrionHDFSSink (#415)
- [FEATURE] Hashing based collections for OrionMongoSink and OrionSTHSink (#420)
- [HARDENING] Set data_model to collection-per-entity (#419)
- [BUG] Use multiple Fiware-ServicePaths in the sinks (#384)
0.8.0
- Ordered death of Cygnus if the logging system fails or it is stoped (#320)
- Management interface port is not opened twice anymore (#302)
- Added a polling interval parameter for reloading the Flume configuration (#334)
- Bug fix, flume-conf.properties.template file from Apache Flume is not copied anymore to Cygnus folder (#348)
- Cygnus exits if the HTTPSource listening port is already in use (#340)
- Added OrionMongoSink (#363)
- Bug fixed: matching table name can be null (no Flume parameter is used) or wrong (#367)
- Add specific sink documents, pointing to them from the README (#322)
- UTC/Z time zone detail added to all the sink timestamps (#370)
- Added OrionSTHSink (#19)
- Multitenancy support in HDFS (#338)
- ISODate objects are used for timestamping in OrionMongoSink (#372)
- Let the origin offset for months in aggregated points starts by 1 instead of 0 (#381)
0.7.1
0.7.0
- Added Kerberos authentication in OrionHDFSSink (https://github.com/telefonicaid/fiware-connectors/issues/290)
- Added posibility to start multiple instances of Cygnus (https://github.com/telefonicaid/fiware-connectors/issues/299)
- Bug fixing: same port in log4j example about two Cygnus instances (https://github.com/telefonicaid/fiware-connectors/issues/305)
- Bug fixing: error in configuration template (cygnusagent.sinks.hdfs-sink.krb5_auth.krb5_login_conf_file) (https://github.com/telefonicaid/fiware-connectors/issues/308)
- Encourage the usage of FQDNs regarding HDFS namenodes when using Kerberos (https://github.com/telefonicaid/fiware-connectors/issues/309)
- Fixed the RoundRoundChannelSelector, now it works for multiple storages (https://github.com/telefonicaid/fiware-connectors/issues/298)
- Allow for infinite TTL (https://github.com/telefonicaid/fiware-connectors/issues/236)
- Added OrionTestSink (https://github.com/telefonicaid/fiware-connectors/issues/307)
- Fixed OrionMySQLSink logger (now it logs as OrionMySQLSink, not as OrionHDFSSink) (https://github.com/telefonicaid/fiware-connectors/issues/321)
0.4
Milestone about the releasing of fiware-connectors 0.4.
Regarding Cygnus, the main features within this release should be:
- New documentation schema based on
doc/
(https://github.com/telefonicaid/fiware-connectors/issues/83) - Smarter channel-based reliability (telefonicaid/fiware-connectors#52)
Regarding the resources, the main content within this release should be:
- Script for removing old data within CKAN-based short term historic (https://github.com/telefonicaid/fiware-connectors/issues/62)
- CKAN basic management script (PR https://github.com/telefonicaid/fiware-connectors/pull/99)