[DO NOT MERGE] Validate beam 2.74.0 RC1#3833
Conversation
Codecov Report❌ Patch coverage is
❌ Your patch check has failed because the patch coverage (0.00%) is below the target coverage (80.00%). You can increase the patch coverage or adjust the target coverage. Additional details and impacted files@@ Coverage Diff @@
## main #3833 +/- ##
============================================
+ Coverage 53.99% 54.95% +0.95%
- Complexity 6895 6923 +28
============================================
Files 1096 1096
Lines 67352 67355 +3
Branches 7558 7558
============================================
+ Hits 36368 37016 +648
+ Misses 28515 27866 -649
- Partials 2469 2473 +4
🚀 New features to boost your workflow:
|
|
I suspect it's some teardown step gets stuck. The last Dataflow job already finished after 2h run the test kick off. |
|
It is com.google.cloud.teleport.templates.DatastoreToTextIT stuck indefinitely. ssh'd into the VM running the test and did a thread dump, suspicious thread is name="testFirestoreToText" classname="com.google.cloud.teleport.templates.DatastoreToTextIT" time="13778.677" and once I run stuck at staging templates for some reason |
|
There are many blocking threads due to some (benign) error happens when the uploaded artifact already exists: and stuck inside This log line is added in GoogleCloudDataproc/hadoop-connectors#1094, due to gcsio 3.x upgrade: apache/beam#38419 |
|
Actually now I think it's an pre-existing bug in TemplateTestBase.java It uses When staging templates, slf4j logs are not emitted: Then gcsio 3.0 added many Google Fluent log, which does not use slf4j, and defaults to JUL's ConsoleLog, suddenly there are many logs in the buffer Testing on https://github.com/GoogleCloudPlatform/DataflowTemplates/actions/runs/26236531476 now |
No description provided.