Hi everyone, below you will find the updates of the Ubuntu Server team members from the last week. If you are interested in discussing a topic please start a thread in the Server area of this Discourse site.
LP: #1823441 SRU for xen FTBFS issue
- Patch is available upstream to fix the bug, and works locally w/ gcc8.
- Packaged patch with fix and uploaded to PPA.
- PPA is building against gcc9, but failing due to other issues.
- Setting bug aside for now until clearer what to do re: gcc9.
LP: #1797836 SRU for logwatch
- Read through logwatch intro guides & man pages, and studied internals to learn how it works more deeply.
- Reproduced error in LXC, uploaded fix to PPA, and submitted MP branch for review.
LP: #1804542 SRU for bind9
- Still blocked waiting on reproduction steps.
LP: #1835968 SRU for Ruby2.5
- For ruby2.5 bug extracted patches separately and reviewed.
- Test case had some implementational errors, but once fixed it reliably reproduced the problem. Used it to test extracted patches, and found the problematic one, which changed how newline handling is done.
- Isolated a one-line fix that reliably addresses the issue without impacting remaining changes.
- Packaged up fix, created PPA for testing, and submitted branch for review.
- Refinement WIP on a few shell scripts to improve error handling and general code tidiness.
- Updated coverity fix from Robie’s comments
- usmerges tool
- Created cron job to run this daily
- pyzmq test timeout fix for andreas. Approved as implemented, but I dug through the code a bit and suggested some additional troubleshooting steps to help identify root cause.
- pacemaker autopkgtest fix for rafael. Approved.
- qemu merge for rafael. Approved.
- terminator fix for annoyingly transparent separators in gtk3 for Tobias Klevenz. Needs Fixing: Addresses the bug but adds some visual glitches; I suggested a fix to be considered to include.
- Working through a bunch of debugging on LP: #1834875
- There is some sort of race happening between udev and the partition table that’s taking a while to nail down, in part because it doesn’t reproduce reliably
- Filed LP: #1838032, which would make lining up the journal entries from other units with cloud-init log entries much easier
- Thanks to Tobias Koch for his invaluable assistance!
- Provided feedback on LP: #1837927 about how to configure ds-identify on a broken OpenStack cloud so that boot will happen when ds-identify would otherwise be in enforcing mode (e.g. Ubuntu 18.04 and later)
- This caused me to file LP: #1838092, because we shouldn’t expect our users to find a shell script in our source tree to work out how to configure their instances
- Landed MP 369783 to make network config source order configurable by data sources
- Landed MP 370526 to split cmdline and initramfs network config determination
- Previously we didn’t distinguish between configuration specified explicitly by the user on the kernel cmdline, and configuration that was passed through to us by the initramfs (in iSCSI boots, for example); this makes those addressable as separate network config sources
- Working on fixing #694
- This is mostly there, just working out how to structure some refactoring that it needs
- Re-reviewed PR #690
- Some extra catch up early in the week due to vacation last week
- Released new jenkins-job-linter snap prompted by an automated security-update-in-the-archive message