[Failing Test]: long running StreamingDataflowWorkerTest

[Abacn]

What happened?

PreCommit Java running time has again elevated. Checking JUnit report found StreamingDataflowWorkerTest takes >1h to complete. Slowest tests are

testExceptionInvalidatesCache[0: [streamingEngine=false]] | 5m3.31s | passed
testExceptionInvalidatesCache[1: [streamingEngine=true]] | 5m3.20s | passed
testLimitOnOutputBundleSizeWithMultipleSinks[0: [streamingEngine=false]]	5m0.94s	passed
testLimitOnOutputBundleSizeWithMultipleSinks[1: [streamingEngine=true]]	5m1.87s	passed
testLimitOnOutputBundleSize[0: [streamingEngine=false]]	5m0.66s	passed
testLimitOnOutputBundleSize[1: [streamingEngine=true]]	5m1.23s	passed
testUnboundedSourceWorkRetry[0: [streamingEngine=false]]	5m6.24s	passed
testUnboundedSourceWorkRetry[1: [streamingEngine=true]]	5m4.58s	passed
testUnboundedSourcesDrain[0: [streamingEngine=false]]	5m2.20s	passed
testUnboundedSourcesDrain[1: [streamingEngine=true]]	5m2.24s	passed
testUnboundedSources[0: [streamingEngine=false]]	5m4.18s	passed
testUnboundedSources[1: [streamingEngine=true]]	5m3.32s	passed

also, in WindmillStateInternalsTest,

testMultimapLazyIterateHugeEntriesResult	1m41.73s	passed
testMultimapLazyIterateHugeEntriesResultSingleEntry	1m40.48s	passed
testMultimapLazyIterateHugeKeysResult	1m24.43s	passed

Issue Failure

Failure: Test is flaky

Issue Priority

Priority: 1 (unhealthy code / failing or flaky postcommit so we cannot be sure the product is healthy)

Issue Components

• Component: Python SDK
• Component: Java SDK
• Component: Go SDK
• Component: Typescript SDK
• Component: IO connector
• Component: Beam YAML
• Component: Beam examples
• Component: Beam playground
• Component: Beam katas
• Component: Website
• Component: Infrastructure
• Component: Spark Runner
• Component: Flink Runner
• Component: Samza Runner
• Component: Twister2 Runner
• Component: Hazelcast Jet Runner
• Component: Google Cloud Dataflow Runner

[Abacn]

Alternatively we can split Dataflow runner as a separate PreCommit

[Abacn]

cc: @acrites @arunpandianp