shunping opened a new pull request, #39663:
URL: https://github.com/apache/beam/pull/39663
### Description
#### Problem
When running Portable/YAML pipelines, gRPC outputs severe error logs
indicating that ManagedChannel instances were garbage collected without being
shut down:
```
SEVERE: *~*~*~ Previous channel ManagedChannelImpl{logId=..., target=...}
was garbage collected without being shut down! ~*~*~*
Make sure to call shutdown()/shutdownNow()
java.lang.Throwable: java.lang.RuntimeException: ManagedChannel allocation
site
at
org.apache.beam.vendor.grpc.v1p69p0.io.grpc.internal.ManagedChannelOrphanWrapper$ManagedChannelReference.<init>(...)
```
#### Cause
Three places in the Java Fn Harness allocated gRPC ManagedChannel instances
without shutting them down upon task completion or harness shutdown:
- ExternalWorkerService: A ManagedChannel was created to query the provision
service for worker options, but was left unclosed after obtaining provisionInfo.
- BeamFnDataGrpcClient & BeamFnDataGrpcMultiplexer: BeamFnDataGrpcClient
created a ManagedChannel per BeamFnDataGrpcMultiplexer, but neither stored nor
shut down the channel when close() was called.
- FnHarness: The control service ManagedChannel was created during harness
startup but was never shut down when FnHarness.main() finished.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
