Start rhsmcertd after sub-man registers in BATS #1899
Conversation
|
This feels wrong. I was reading the sub-man diff as "will send profile itself if rhsmcertd is not running", is that part not working? |
roles/bats/defaults/main.yml
Outdated
| @@ -17,7 +17,6 @@ bats_tests: | |||
| - "fb-test-foreman.bats" | |||
| - "fb-katello-content.bats" | |||
| - "fb-katello-container.bats" | |||
| - "fb-katello-client-katello-agent.bats" | |||
There was a problem hiding this comment.
Can we keep that for now -- Satellite 6.14 is still supported ;)
There was a problem hiding this comment.
Ah I didn't realize we reused the same BATS for such old Foremans.
If it's about rhsmcertd then candlepin/subscription-manager@a8acd2d looks like a relevant change. Is there anything in the journal for |
|
It is relevant, yes, we're talking to sub-man developers about this right now :) |
ianballou
force-pushed
the
rhsmcertd-live-bats
branch
from
307aa40 to
552b534
Compare
|
Question is -- should we merge this because it likely doesn't hurt anything and will likely get the pipeline running again until we have our fix? |
ianballou
force-pushed
the
rhsmcertd-live-bats
branch
2 times, most recently
from
86818d1 to
8d26e5a
Compare
|
The gist of the issue is that sub-man is no longer sending the package profile when rhsmcertd is dead, but it should be. |
|
Yeah, I am cool merging this while we wait for the subman fix |
|
I have a general rule to at least refer to an issue in a comment to track why we have workarounds and make it clear it is a workaround. |
ianballou
force-pushed
the
rhsmcertd-live-bats
branch
from
8d26e5a to
020ea42
Compare
|
I couldn't find an issue so I just used subscription-manager 1.29.46 as the point of removal. |
subscription-manager is no longer sending package profiles without rhsmcertd as of
1.29.45.By default, the service is dead until the system is rebooted. This should theoretically fix the current failing errata tests.
Also, I deleted the katello-agent tests since it's very gone by now.