veblush commented on pull request #11651:
URL: https://github.com/apache/beam/pull/11651#issuecomment-631882467
Thanks, Brian and Chamikara!
This is an automated message from the Apache Git Service.
To respond to the message,
veblush commented on pull request #11651:
URL: https://github.com/apache/beam/pull/11651#issuecomment-631815316
Thanks for running the test and the result doesn't make sense because this
change won't introduce any runtime behavior other than dependency. Is it
possible to be caused by flaky
veblush commented on pull request #11651:
URL: https://github.com/apache/beam/pull/11651#issuecomment-631632732
This is the result of linkage chekcer. (it was challenging because it seems
to require Java8 strictly to finish it smoothly)
```
# /bin/bash sdks/java/build-tools/beam-
veblush commented on pull request #11651:
URL: https://github.com/apache/beam/pull/11651#issuecomment-631065488
It turns out that `grpc-netty` and `grpc-netty-shaded` can coexist. So I'm
trying to list those new dependencies in the build file to make gRPC component
have the same version.
veblush commented on pull request #11651:
URL: https://github.com/apache/beam/pull/11651#issuecomment-631025815
This is most likely caused by gcsio 2.1.3 dependency to `grpc-netty-shaded`
([ref](https://mvnrepository.com/artifact/com.google.cloud.bigdataoss/gcsio/2.1.3)).
Meanwhile, Beam u
veblush commented on pull request #11651:
URL: https://github.com/apache/beam/pull/11651#issuecomment-631024519
It appears that it actually has a problem with dependency.
Java PostCommit /
[testReadEmptySketchFromBigQuery](https://builds.apache.org/job/beam_PostCommit_Java_PR/3
veblush commented on pull request #11651:
URL: https://github.com/apache/beam/pull/11651#issuecomment-630948467
No performance regression with this change. (Tested with internal tests)
This is an automated message from the Ap