action #70843
closed
OSD automatic deployment failed in Sept, 02, 2020
Added by Xiaojing_liu over 3 years ago.
Updated over 3 years ago.
Description
The reason of deployment failure is: https://gitlab.suse.de/openqa/osd-deployment/-/jobs/252441#L409
( 1/12) Installing: kernel-default-5.8.5-2.1.g34111a4.aarch64 [...error]
Installation of kernel-default-5.8.5-2.1.g34111a4.aarch64 failed:
Error: Subprocess failed. Error: RPM failed: error: unpacking of archive failed: cpio: Bad magic
error: kernel-default-5.8.5-2.1.g34111a4.aarch64: install failed
- Status changed from New to In Progress
- Assignee set to okurz
- Target version set to Ready
thank you for picking that up and working on it so far. From chat I read that you downloaded kernel-default manually and tried to convert manually to cpio. I do not think that should be necessary. Maybe something in rpm cache needs handling.
find /var/cache/ -name '*kernel-default*' -delete ; zypper in kernel-default does not work either. I assume the kernel stable repo is currently not usable after the switch in Factory to use a different rpm format, that's what I read somewhere on a mailing list. As a workaround I did zypper al kernel-default so now zypper dup does not try to upgrade the kernel anymore. I retriggered the deployment CI step with https://gitlab.suse.de/openqa/osd-deployment/-/jobs/252627
- Status changed from In Progress to Feedback
- Priority changed from High to Low
- Status changed from Feedback to Resolved
openqaworker-arm-2 and openqaworker-arm-3 are back. I have not found any information about the kernel repo or how to be able to use it. As the newer kernel does not show an obvious improvement over the old stable maintained one I have opted instead to disable the repo "Kernel_stable" and removed the lock and applied an upgrade to use the old stable maintained kernel instead on openqaworker-arm-1 and openqaworker-arm-2 (openqaworker-arm-3 is still using old stable maintained kernel anyway):
zypper mr -d 2
zypper rl kernel-default
zypper dup
After this turns out clean I can restart the deployment job from today which failed for the same reason as above -> https://gitlab.suse.de/openqa/osd-deployment/-/jobs/255282 turned out to be fine. deployment completed. I consider this ticket resolved regardless of the change of the kernel repo as there is no easy solution known to me and kernel head does not seem to provide any significant benefit.
Also available in: Atom
PDF