Right now, bin/pungi-wait-for-signed-ostree-handler has:
count = 0 while not os.path.exists(path): ts_log("Commit not signed yet, waiting...") count += 1 if count >= 60: # Repeat every 5 minutes print('Repeating notification') fedmsg.publish(topic=topic, modname='pungi', msg=data) count = 0 time.sleep(5) print("Found signature, waiting for ref to be updated.") ref_file = os.path.join(repo, "refs/heads", data["ref"]) while not is_ref_updated(ref_file, commit): ts_log("Ref is not yet up-to-date, waiting...") time.sleep(5) print("Ref is up-to-date. All done!")
However, when autosign signs (and the commitmeta exists), but errors updating the ref, pungi just sits waiting for that ref forever. It might be nice to resend the fedmsg for the ref case as well, as autosign currently can get in a state where it writes commitmeta, but fails to update the ref and never tries again.
So, adding a publish loop to the ref case, or putting both the commitmeta/ref case inside the publish loop would be welcome.
Commit 16d6e5b0 fixes this issue
Metadata Update from @lsedlar: - Issue set to the milestone: 4.2.1
This issue has been migrated to Fedora Forge: https://forge.fedoraproject.org/pungi/pungi/issues/1350
Please continue any further discussion there.