#3554 side tag with builds was removed
Closed: Invalid by tkopecek. Opened by zdohnal.

Hi all,

I'm filing this issue based on the communication on devel mailing thread https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/message/GSVWZDESIQ7GSBPLLWWT7QGTKRC6ZETB/ - I had a side tag, which contained several builds, but it got removed after two weeks.

The side tag name was f38-build-side-58658, created on the 22nd September and it was no longer available on 9th October (slightly after 14 days).

Would you mind looking into it?

Thank you in advance!

Zdenek


Hi, can you please reopen this in https://pagure.io/releng/issues ? This is tracker for koji development issues, while yours is about Fedora's instance which is handled by Fedora relengs.

Metadata Update from @tkopecek:
- Custom field Size adjusted to None
- Issue close_status updated to: Invalid
- Issue status updated to: Closed (was: Open)

Well, I think this is a koji upstream issue. :)

We run:

"/usr/sbin/koji-sidetag-cleanup --empty-delay=14 --old-delay=30"

This sidetag ( f38-build-side-58658 ) had a number of tagged / active builds in it, it wasn't empty, but it was removed after 14 days.

So, it seems like empty-delay is really doing inactive delay?

Metadata Update from @kevin:
- Issue status updated to: Open (was: Closed)

Ah, probably found. It was probably deleted by "inactive" cleaner which has default of 10 days. If you don't want to utilize this cleanup method (just delete anything what wasn't touched for some time), add --no-inactive or modify --inactive-delay

$ koji -p fedora list-history --tag f38-build-side-58658 -e
Thu Sep 22 16:00:47 2022 (eid 99802843) new build target: f38-build-side-58658 by zdohnal
Thu Sep 22 16:00:47 2022 (eid 99802843) new tag: f38-build-side-58658 by zdohnal
Thu Sep 22 16:00:47 2022 (eid 99802843) added tag option sidetag for tag f38-build-side-58658 by zdohnal
Thu Sep 22 16:00:47 2022 (eid 99802843) added tag option sidetag_user for tag f38-build-side-58658 by zdohnal
Thu Sep 22 16:00:47 2022 (eid 99802843) added tag option sidetag_user_id for tag f38-build-side-58658 by zdohnal
Thu Sep 22 16:00:47 2022 (eid 99802843) inheritance line f38-build-side-58658->f38-build added by zdohnal
Thu Sep 22 17:11:28 2022 (eid 99804171) qpdf-11.1.0-1.fc38 tagged into f38-build-side-58658 by zdohnal
Thu Sep 22 17:28:53 2022 (eid 99804409) cups-filters-1.28.16-4.fc38 tagged into f38-build-side-58658 by zdohnal
Sat Sep 24 03:20:58 2022 (eid 99830871) python-pikepdf-6.0.2-1.fc38 tagged into f38-build-side-58658 by qulogic
Sun Sep 25 17:56:02 2022 (eid 99873639) pdfmixtool-1.1.1-1.fc38 tagged into f38-build-side-58658 by xfgusta
Thu Oct  6 06:00:18 2022 (eid 100323717) build target deleted: f38-build-side-58658 by oscar
Thu Oct  6 06:00:18 2022 (eid 100323717) tag deleted: f38-build-side-58658 by oscar
Thu Oct  6 06:00:18 2022 (eid 100323717) tag option sidetag removed for f38-build-side-58658 by oscar
Thu Oct  6 06:00:18 2022 (eid 100323717) tag option sidetag_user removed for f38-build-side-58658 by oscar
Thu Oct  6 06:00:18 2022 (eid 100323717) tag option sidetag_user_id removed for f38-build-side-58658 by oscar
Thu Oct  6 06:00:18 2022 (eid 100323717) inheritance line f38-build-side-58658->f38-build removed by oscar
Thu Oct  6 06:00:18 2022 (eid 100323717) cups-filters-1.28.16-4.fc38 untagged from f38-build-side-58658 by oscar
Thu Oct  6 06:00:18 2022 (eid 100323717) qpdf-11.1.0-1.fc38 untagged from f38-build-side-58658 by oscar
Thu Oct  6 06:00:18 2022 (eid 100323717) python-pikepdf-6.0.2-1.fc38 untagged from f38-build-side-58658 by oscar
Thu Oct  6 06:00:18 2022 (eid 100323717) pdfmixtool-1.1.1-1.fc38 untagged from f38-build-side-58658 by oscar

Ah ha, that makes sense. It's a bit unexpected to have a new set of defaults appear, but I guess we should have been paying more attention...

Agreed, I shouldn't have done it this way.

Metadata Update from @tkopecek:
- Issue close_status updated to: Invalid
- Issue status updated to: Closed (was: Open)

This issue has been migrated to Fedora Forge:
https://forge.fedoraproject.org/koji/koji/issues/3554

Please continue any further discussion there.

Metadata