Comments (10)
The failing tests are
2024-06-06T15:49:25.5514012Z teststream_test.go:82: Failed to execute job: job go0testteststreamsimple0infinitydefault0645-runner-0606154852-104bc19c_903ad8b0-583f-49fa-9c25-7de441e9645e failed:
2024-06-06T15:49:25.5516578Z org.apache.beam.sdk.coders.CoderException: `UnknownCoderWrapper` was used to perform an actual decoding in the Java SDK. Potentially a Java transform is being followed by a cross-language transform that uses a coder that is not available in the Java SDK. Please make sure that Python transforms at the multi-language boundary use Beam portable coders.
2024-06-06T15:49:25.5518501Z --- FAIL: TestTestStreamSimple_InfinityDefault (36.04s)
2024-06-06T15:48:49.4744385Z teststream_test.go:77: Failed to execute job: job go0testteststreamsimple0300-runner-0606154839-3044966_f7c93228-67e5-4745-8d98-66cf379629eb failed:
2024-06-06T15:48:49.4748151Z org.apache.beam.sdk.coders.CoderException: `UnknownCoderWrapper` was used to perform an actual decoding in the Java SDK. Potentially a Java transform is being followed by a cross-language transform that uses a coder that is not available in the Java SDK. Please make sure that Python transforms at the multi-language boundary use Beam portable coders.
2024-06-06T15:48:49.4751550Z --- FAIL: TestTestStreamSimple (13.58s)
2024-06-06T15:49:38.9044506Z teststream_test.go:87: Failed to execute job: job go0testteststreamtogbk02-runner-0606154928-bee4a43d_deed8bb6-6a4b-42bf-9d42-2bd1911d0bfe failed:
2024-06-06T15:49:38.9048296Z org.apache.beam.sdk.coders.CoderException: `UnknownCoderWrapper` was used to perform an actual decoding in the Java SDK. Potentially a Java transform is being followed by a cross-language transform that uses a coder that is not available in the Java SDK. Please make sure that Python transforms at the multi-language boundary use Beam portable coders.
2024-06-06T15:49:38.9050512Z --- FAIL: TestTestStreamToGBK (13.42s)
2024-06-06T15:49:48.8207809Z teststream_test.go:92: Failed to execute job: job go0testteststreamtimerseventtime055-runner-0606154942-51c50c75_ae30c901-6f2f-4981-a593-956e0a3a0b79 failed:
2024-06-06T15:49:48.8209479Z java.lang.NullPointerException
2024-06-06T15:49:48.8210362Z --- FAIL: TestTestStreamTimersEventTime (9.83s)
These tests were added in #31046 . Shall we fix the test or disable them on Flink VR test suites? @lostluck
from beam.
cc: @kennknowles another currently permared PostCommit on release-2.57.0 branch
from beam.
The error also sounds like #30994. The error message notes "Python transforms" which is not accurate either
from beam.
Agreed, that filtering is the right move here. Those tests do use strings which is one of the affected datatypes Flink's TestStream corrupts.
At some point we determined that Flink does with TestStream and mutates those coders (adding length prefixes where they weren't previously) without making the equivalent mutations to the equivalent bytes, but we weren't able to pin down where it was coming from.
This we filtered out those tests here:
https://github.com/apache/beam/blob/master/sdks/go/test/integration/integration.go#L181
Ideally we fix the Flink test stream implementation, but until then, we filter, since it's not commonly used.
from beam.
Actually, "TestTestStreamSimple" and "TestTestStreamToGBK" should be working, so those are new failures. "TestTestStreamTimersEventTime" I'd expect to fail based on previous behaviour.
from beam.
Actually, "TestTestStreamSimple" and "TestTestStreamToGBK" should be working
All these 4 tests were added in #31046 and failing since first run. Or do you suggest the newly added test reveals some underlying bug/gap ?
from beam.
Ah! Right I recall now. So those were added because they did reveal a gap in Prism's test stream implementation.
They're likely revealing one in Flink, so agreed they should be filtered.
The simple ones are pipelines without any Impulse transform, so the runner's TestStream must be capable of kicking off the pipeline.
from beam.
so agreed they should be filtered.
yeah, thanks, I am trying to do that. However it seems not obvious how can I filter out a specific test for a specific runner in VR test suite.
In Java this was done by excludeCategories https://github.com/apache/beam/blob/master/runners/flink/flink_runner.gradle#L298 (as gradle is built for java)
from beam.
We filter out those tests flink here:
https://github.com/apache/beam/blob/master/sdks/go/test/integration/integration.go#L181
Each runner has its own list of tests it can't run.
Wildcards or the raw test name can be used too.
from beam.
Ah, this is great news. I had not dug very deeply into the failures. The whole branch looked a mess so I assumed that our GHA just was broken in a big way.
from beam.
Related Issues (20)
- [Bug]: KafkaIO Should not throw errors for empty topics.
- [Bug]: SolaceIO.Read: clients are not properly closed. HOT 1
- [Task][Prism]: Create a PrismRunnerRegistrar
- The StressTests Java BigQueryIO job is flaky
- [Bug]: Beam YAML provider parsing breakage
- [prism] Metrics from runner expanded composites aren't mapped back to user terms.
- [Bug]: CoGroupByKey throws with large (10K+) output groups on Flink runner HOT 2
- [Failing Test]: Python BigQuery Test - PipelineBasedStreamingInsertTest::tes_batch_size_with_auto_sharding_0 is flaky
- [Feature Request]: add support for Datastore Database ID in Python SDK
- [prism] Java PipelineTest: Unknown Urn "ProjectTag", "IdentityTransform" - Support "Empty" Composites without subtransforms.
- [prism] Preprocess failure - Expected Runner Flatten Node - but wasn't
- [Bug]: Java SDK BigQueryIO's RowMutationInformation class is not backward compatible with previous releases
- [CsvIO]: Update class for representing date-times in CsvIOParseHelpers::parseCell(String, Schema.Field)
- [Failing Test]: testRedistributeAfterSlidingWindows incorrect HOT 2
- [Bug]: Failed to restore state on Flink runner HOT 1
- [Bug]: Portable runner queryMetrics returns duplicate result
- [prism] Support Empty Flattens as Side Input - no way to make progress with pending elements
- [prism] Java and Python SplittableDoFnTests - invalid stream header (likely coders)
- [CsvIO]: Add CsvIORecordToObjects Class
- Performance Regression or Improvement: pytorch_image_classification_benchmarks-resnet101-mean_load_model_latency_milli_secs:mean_load_model_latency_milli_secs
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from beam.