Closed github-actions[bot] closed 5 months ago
Monday
Checking buildfarm healthiness Overall looks OK, gazebo farm is getting higher demand on a specific node tagged for debbuilders given that debbuilders in noble are processed now as well, checking with Gazebo team if this is expected (I think it is)
Stopped hanging job, 3 days and 9 hours: https://ci.ros2.org/job/nightly_win_rel/2992/ Parallel executor does not give us relevant info about the hanged job in the logs :/ it seems it only reports once the test suite ends.
[INVESTIGATION]
Linter errors related to point_cloud_transport: Reference builds: https://ci.ros2.org/view/nightly/job/nightly_linux-aarch64_debug/2715/ https://build.ros2.org/view/Rci/job/Rci__nightly-release_ubuntu_jammy_amd64/783/
Reported to the ROS 2 team to check if someone is looking at it, should be an easy fix. Not sure if it's related to flake8 transitions.
[INVESTIGATION] Internal tools failing, see: https://github.com/osrf/buildfarm-tools/actions/runs/8401493877/job/23009932079
Of course, Cristobal is on PTO so some things gotta break :laughing: Taking a look to the problem, it seems it's related to an annotation not being recognized by the script for some reason.
Ugh, is not related to Python, is related to Ruby.
This ruby script: https://github.com/osrf/buildfarm-tools/blob/47eaf84e2cd773f4f0d5b09bea7b9d2d9c6be07a/database/scripts/check_buildfarm.rb#L42 to be specific.
This error broke it: point_cloud_transport.flake8.E261 (./doc/conf.py:37:21)
I think the script doesn't like the spaces, probably requires modifying: https://github.com/osrf/buildfarm-tools-private/blob/c69a4256413713d9bc7faf8e00d972cabe504738/databaseFetcher.py#L89
Debugging:
puts %x{echo calculate_flakiness_jobs.sql '#{tr['error_name']}' "15 days"}
And this prints
calculate_flakiness_jobs.sql point_cloud_transport.flake8.E261 (.\doc
Some PRs were merged yesterday that broke it. Chris is addressing it today. The combination of these PRs should fix the issue:
Seems like an issue with Fastdds it has been failing since build #83 on Feb 16th.
/tmp/ws2/build_isolated/performance_test/src/idlgen/fast_rtps/gen/fast_rtps/Header_PubSubTypes.cxx:23:10: fatal error: fastdds/rtps/common/CdrSerialization.hpp: No such file or directory
Might need the fix implemented for humble, Chris will take a look at it.
SERVERS and CAPACITY OK
Odd build regression on Rci_nightly_Release yesterday: https://build.ros2.org/view/Rci/job/Rci__nightly-release_ubuntu_jammy_amd64/785/ -> Keeping track to see if it fails again
Odd build regression on doc_rosindex yesterday: https://build.ros.org/job/doc_rosindex/1937/ -> Keeping track to see if it fails again
Chris fixing Windows_debug: https://osrf.slack.com/archives/C04R65U5CNA/p1711542127941529 (Keeping track)
Reference jobs: https://ci.ros2.org/job/nightly_linux_iron_coverage/378/ https://ci.ros2.org/job/nightly_linux_humble_coverage/751/
Probably should be solved with backports, raised to ROS 2 team on: https://osrf.slack.com/archives/C04R65U5CNA/p1711550109739249
Previous log #34