Skip to content

Epsio Engine

20.12.3

Minimum K8S operator version required: v1.0.2

Bug Fixes:

  • Fix high availability not running when there are database connectivity issues.

20.12.2

Minimum K8S operator version required: v1.0.2

Bug Fixes:

  • Fix issue with sinking into tables with unique constraints.

20.12.1

Minimum K8S operator version required: v1.0.2

Bug Fixes:

  • Fix new install script not running on some machines.

20.12.0

Minimum K8S operator version required: v1.0.2

Enhancements:

  • Support json_agg and array_agg on jsons.
  • Create more optimized DELETE query when running above a partitioned table.

20.11.0

Minimum K8S operator version required: v1.0.2

Enhancements:

  • Add initial support for HA Epsio deployments.

Bug Fixes:

  • Clean responses table after every command.
  • Fix race condition causing CDC to drop some relevant diffs on restart.
  • Add better support for catching MySQL TRUNCATE commands.
  • Fix MySQL error handling zero dates.
  • Fix issue with cursor name in PostgreSQL when creating views with long base table names.

20.10.0

Minimum K8S operator version required: v1.0.2

Enhancements:

  • Add significant performance and memory improvements.
  • Add regex expressions support (LIKE/ILIKE).
  • Improve state memory usage.
  • Add jsonb aggregation support.

Bug Fixes:

  • Fix memory issues with sinker.
  • Fix unreadable table validation errors.

20.9.0

Minimum K8S operator version required: v1.0.2

Enhancements:

  • Add support for MSSQL.
  • Add support for RANK/ROW_NUMBER window functions.
  • Add support for lower and upper functions.
  • Add support for views backup table in MySQL.
  • Restart diffs stats after view restart.
  • Add date to timestamp coercion support.

Bug Fixes:

  • Fix bug with long errors messages in MySQL's list_views.
  • Fix MySQL forwarder error getting stuck if Binlog was discarded.

20.3.1

Minimum K8S operator version required: v1.0.2

Bug Fixes

  • Fix issue causing Datadog metrics to not be sent after 19.x version migration

20.3.0

Minimum K8S operator version required: v1.0.2

Enhancements:

  • Major improvements in memory footprint.
  • New restart & auto-restart mechanism.
  • PostgreSQL wraparound handling without view restart.
  • Support for JSON output in docker logs.
  • Epsio can now resume a view without a restart after a sink error (partition missing, etc).
  • Epsio now survives ALTER of columns without restart.
  • ENUM support in PostgreSQL.
  • New version upgrade migration mechanism.
  • Performance improvements

Bug Fixes

  • Restart now "survives" pod restart.
  • Fix issue causing Epsio to calculate cached memory as used memory.
  • Fix MySQL bug during installation process.

19.3.12

Enhancements:

  • Performance boost during diffs consolidation.

19.3.11

Bug Fixes:

  • Fix additional potential race conditions during PostgreSQL failover in some PostgreSQL versions.

19.3.10

Bug Fixes:

  • Fix potential race conditions during PostgreSQL failover in some PostgreSQL versions.

19.3.8

Enhancements:

  • Remove unnecessary logs.

19.3.7

Bug Fixes:

  • Fix bug causing tables to remain after drop_view, if the view was previously restarted/used an existing table.

19.3.6

Bug Fixes:

  • Fix bug in latency calculation.

19.3.3

Bug Fixes:

  • Fix issues with invalid consolidation of latencies.

19.3.2

Enhancements:

  • Performance optimization in tracing.

19.3.0

Bug Fixes:

  • Fixed issue causing autorestart to restart the same view multiple times.
  • Fixed MySQL collation issue in alter_view.

19.2.0

Enhancements:

  • create_view on an existing table now synchronizes previous data from existing table and updates only new data (performance boost during create_view)

19.1.1

Bug Fixes:

  • Hotfix for aggregations utilizing Datetime with timezone.

19.1.0

Enhancements:

  • Send disk usage stats in datadog.

Bug Fixes:

  • Fix issues with log files becoming too large when disk space is low.

19.0.0

Enhancements:

  • Add support for numerics with arbitrary precisions in PostgreSQL.

Bug Fixes:

  • Fix issues when dumping traces and logs.

18.1.0

Enhancements:

  • Added BIT type support in MySQL.

18.0.2

Bug fixes:

  • Fixed issue causing views to freeze when selecting a constant.

18.0.0

Enchantments:

  • Support various collations in MySQL.

Bug Fixes:

  • Fixed issue regarding auto view snapshot restore.

17.5.1

Enhancements:

  • Added consolidation of diffs before JOIN operations (performance boost).

17.5.0

Bug Fixes:

  • Fixed DataDog integration, which was sending bad service checks when using the same token for multiple deployments.

17.4.0

Enhancements:

  • Added performance improvements in CDC.

Bug Fixes:

  • Fixed issue when selecting constants without any tables in the query.

17.3.0

Enhancements:

  • Changed the method of populating the temp table in PostgreSQL to be an unlogged table, improving performance.

Bug Fixes:

  • Improved wraparound handling in PostgreSQL.
  • Fixed memory issues that occurred when writing to the sink.

17.2.0

Enhancements:

  • Add Datadog integration for metrics / alerting.
  • Enhancements in Terraform deployment.
  • Improved memory footprint on small deployments.
  • Add support for EPSIO_ prefix to docker enviroment variables.
  • Add support for EPSIO_DATABASE_URL env in docker.
  • Add restore_views procedure to restore snapshot.
  • Add EPSIO_SHOULD_RESTORE_FROM_BACKUP_ON_STARTUP=1 env to docker to re-build views from backup on each startup.

Bugfixes:

  • Fixed issue in views with unique constraints
  • Fixed timeout issue with low statement_timeout.
  • Fixed time zone synchronization in list_views when client timezone is different than server timezone.

17.1.3

Enhancements:

  • Expose internal k8s operator

Bugfixes:

  • Fix issue with logging query with many subqueries

17.0.0

Enhancements:

  • Add more statistics for query planning
  • Optimize memory usage in data-flows

Bug Fixes

  • Fix latency calculation in MySQL
  • Fix evaluation of adding interval of seconds

16.3.4

Bug Fixes:

  • Make minimal batch size correlative to system memory for small deployments

16.3.3

Enhancements:

  • Optimizations for planning
  • Optimizations for population
  • Sharing data across operators

16.3.2

Enhancements:

  • Enhanced population speeds in large queries
  • Added better error handling for multi-sources

16.3.1

Bug Fixes:

  • Fixed de-serialization error in cached plans causing re-planning of queries.

16.3.0

Enhancements:

  • Added support for now()
  • Added multiple casting options
  • Added restart_view procedure

Bug Fixes:

  • Fixed statistical bug in large floats