[GH] (pekko-persistence-cassandra/resource-leak-test-actor-systems): Workflow run "Unit Tests" is working again!

[GitBox <gi...@apache.org>]

The GitHub Actions job "Unit Tests" on pekko-persistence-cassandra.git/resource-leak-test-actor-systems has succeeded.
Run started by GitHub user pjfanning (triggered by pjfanning).

Head commit for run:
16a82724f5cf1ac8303b2653e233590133c2a50e / PJ Fanning <pj...@users.noreply.github.com>
Terminate the ActorSystems that specs leave running

Motivation:
PubSubThrottlerSpec and ReconnectSpec create an ActorSystem and never terminate
it, so it lives for the rest of the test JVM run. ReconnectSpec is the worse of
the two, its driver is configured to keep reconnecting, so its threads carry on
after the suite. AbstractEventsByTagMigrationSpec shuts down systemTwo and
systemThree but not migrationSystem.

Modification:
Shut the ActorSystem down in afterAll in all three specs. PubSubThrottlerSpec
verifies the shutdown, ReconnectSpec does not because its driver is left
reconnecting to a Cassandra that has been stopped.

Result:
The suites release their ActorSystems instead of leaving them running for the
rest of the test JVM.

Tests:
- sbt "core/Test/testOnly org.apache.pekko.persistence.cassandra.journal.PubSubThrottlerSpec" - passed, including the verified shutdown
- sbt "core/Test/compile" - passed
- scalafmt --mode diff-ref=upstream/main --list - no changes needed
- ReconnectSpec and EventsByTagMigrationSpec not run - need Cassandra/docker, not available locally

References:
None - resource leak found while auditing actor and session lifecycles

Report URL: https://github.com/apache/pekko-persistence-cassandra/actions/runs/34058872458

With regards,
GitHub Actions via GitBox


---------------------------------------------------------------------
To unsubscribe, e-mail: notifications-unsubscribe@pekko.apache.org
For additional commands, e-mail: notifications-help@pekko.apache.org