Page MenuHomeSoftware Heritage

debian/gpg-buildpackage: Add cassandra repository information
ClosedPublic

Authored by ardumont on Feb 3 2020, 2:53 PM.

Details

Summary

To allow build of the new swh-storage backend implementation.

Test Plan

tox

Diff Detail

Repository
rCJSWH Jenkins jobs
Branch
master
Lint
No Linters Available
Unit
No Unit Test Coverage
Build Status
Buildable 10421
Build 15512: arc lint + arc unit

Event Timeline

This is a very large key, I assume because it contains lots of signatures. They could be removed if the size causes issues.

This revision is now accepted and ready to land.Feb 3 2020, 2:55 PM
jobs/templates/debian/gbp-buildpackage.groovy
3284

mmm, i see more public key block than 1 ...

Only provide one public key block

Unsure whether we can provide more than 1 block.

Turns out that now we have issues because we don't have enough keys

16:04 <+ardumont> https://jenkins.softwareheritage.org/view/Debian%20packages/job/debian/job/packages/job/DSTO/job/gbp-buildpackage/162/console
16:04 <+ardumont> it claims to not find the cassandra repository key file (thus failing to verify
16:04 <+ardumont> the apt repo (apparently the source of the build failure)
16:04 <+ardumont> but the workspace (for that same build) shows there is one cassandra.asc file
16:04 <+ardumont> so there is a file (don't know if it's the right one, mind you, but there is one)...
16:04 <+ardumont> https://jenkins.softwareheritage.org/view/Debian%20packages/job/debian/job/packages/job/DSTO/job/gbp-buildpackage/162/execution/node/3/ws/keys/
16:04 <+ardumont> so i don't get it...
16:04 <+ardumont> also i don't see much movement over our jenkins jobs declarations either, so ¯\_(ツ)_/¯
16:04 <+ardumont> don't get why all of a sudden, it fails (fsvo "sudden", it's been 14 days in between builds after all)
16:04 <+douardda> olasd: so you are handling the azure mess^Wversion update or should I take care of it?
16:05 <+olasd> douardda: I'm taking care of it
16:05 <+douardda> thank you
16:07 <+olasd> ardumont: the cassandra repository signing key is actually a _keyring_ of all the developers
16:07 <+olasd> in 873493d4591b2067a2fa907700bcbf62c2281236, you removed all but one key
16:07 -- Notice(swhbot): swh-jenkins-jobs/873493d459 (author: ardumont) gbp-buildpackage: Use smaller gpg public key <https://forge.softwareheritage.org/rCJSWH873493d459>
16:07 <+olasd> I guess another developer did the last upload
16:09 <+ardumont> ah that would explain it
16:09 <+olasd> the way these keyrings are handled is pretty awful
16:10 <+olasd> well, now that this stuff is jinja2 we could include the keyring from another file
16:10 <+ardumont> i don't recall why i initially did that though (too bad i did not say...)
16:10 <+ardumont> yes, i was gonna ask
16:10 <+olasd> "because the file was large" is my guess
16:10 <+ardumont> i saw some !include-ninja something somewhere
16:12 <+ardumont> thx, i'll fix that ;)

:D