0.3.3

A patch release that bumps dependencies for potential security vulnerabilities.

• Bump apache commons-compress to 1.21 (#76)
• Bump netty-codec to 4.1.68 (#76)
• Bump protobuf-java to 3.16.1 (#76)
• Bump scala version to 2.13.9 (#76)
• Bump gson to 2.8.9 (#76)
• Exclude kafka-clients library (#76)
• Bump postgresql driver to 42.4.1 (#76)
• Bump jackson-databind to 2.12.6.1 (#76)

Version 0.3.2

Enhancements

Checkpoint outstanding records when loader terminates (#69)
Use direct executor for KPL callback (#73)
Warn about bad rows even in DummyStream sink (#64)
Disable colored logging by default (#63)

Bug fixes

Fix flaky sink test: return columns in order (#71)
Fix timestamp in parsing error (#68)
Use ContextShift after Async callback in KinesisSink (#67)

Under the hood

Add latest tag to Docker build explicitly (#61)
Github action to scan docker image with lacework (#72)

Version 0.3.1

Bugfix release.

• Make database thread pool size configurable (#60)

Version 0.3.0

In this release, we are adding new features, fixing some bugs, and bumping a few dependencies.

Upgrading from 0.2.0

Changes on config

We made some changes to the general config structure to make it more in line with other Snowplow products. You can find the description of the new structure here. Also, you can find example configs here.

Changing some of the column types

The database schemas created with the previous version isn't working well with events enriched with pii pseudoanonymization. Therefore, we needed to change types of the some of the columns. You need to apply this migration script to your database.

New features

• Use a local directory as an event source (#42)
• Send bad rows to separate stream (#54)

Improvements

• Improve checkpointing of kinesis records (#39)

Configuration changes

• Make database max pool size configurable (#56)

Bug fixes

• Pass stream name and application name to fanout config object (#44)
• Prepare schema definition to work with pii pseudoanonymization (#45)
• Run tests of different modules seperately in CI/CD (#58)

Dependency upgrades

• Bump sbt to 1.5.5 (#46)
• Bump postgresql to 42.2.23 (#47)
• Bump logback to 1.2.5 (#48)
• Bump decline to 2.1.0 (#49)
• Bump cats-effect to 2.5.3 (#50)
• Bump fs2 to 2.5.9 (#51)
• Bump snowplow-badrows to 2.1.1 (#52)
• Bump schema-ddl to 0.14.1 (#53)

Version 0.2.0

New features

Add ability to disable CloudWatch metric reportings (#11)
Remove requirement to always use Kinesis Enhanced Fan Out (#14)

Bug fixes

Fix type derivation of enum fields (#12)
Ignore missing comment on atomic events table (#36)
Use jsonb instead of varchar for complex enums (#31)

Configuration changes

Improve configuration defaults and examples (#41)
Remove unused name and id fields from config hocon (#38)
Add PubSub configuration to the example config file (#34)
Rename storage to output in config file (#33)
Use hocon for cli config instead of self-describing json (#20)

Dependency upgrades

Bump postgresql to 42.2.0 (#15)
Bump schema-ddl to 0.13.1 (#27)
Bump log4s to 1.10.0 (#26)
Bump fs2 to 2.5.6 (#25)
Bump doobie to 0.13.4 (#24)
Bump fs2-google-pubsub to 0.17.0 (#23)
Bump fs2 aws to 3.1.1 (#22)
Bump circe to 0.14.1 (#30)
Bump cats-effect to 2.5.1 (#29)
Bump decline to 2.0.0 (#28)
Bump commons-codec to 1.15 (#21)
Use AdoptOpenJDK 11 as docker base image (#18)

Under the hood

Build docker images for amd64 and arm64 architectures (#19)
Publish to maven central instead of bintray (#17)
Attach jar files to github release (#10)

0.2.0-rc1

Use jsonb instead of varchar for complex enums (close #31)

0.1.1

Bugfix release.

• Add support for nullable fields (#6)
• Bump snowplow-scala-analytics-sdk to 2.1.0 (#8)

0.1.0

Initial release.