Couple of my updates have e-mailed me $subj. Is it something to worry about?
Example: https://bodhi.fedoraproject.org/updates/FEDORA-2019-f2ffe568a1
On Sun, Oct 6, 2019, 15:07 Miro Hrončok mhroncok@redhat.com wrote:
Couple of my updates have e-mailed me $subj. Is it something to worry about?
I got this too for a lot of my updates, just a few hours ago. I assumed it was caused by some kind of server glitch, maybe related to the current koji outage.
Fabio
Example: https://bodhi.fedoraproject.org/updates/FEDORA-2019-f2ffe568a1
Miro Hrončok
Phone: +420777974800 IRC: mhroncok _______________________________________________ devel mailing list -- devel@lists.fedoraproject.org To unsubscribe send an email to devel-leave@lists.fedoraproject.org Fedora Code of Conduct: https://docs.fedoraproject.org/en-US/project/code-of-conduct/ List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org
On Sun, Oct 6, 2019 at 1:12 PM Fabio Valentini decathorpe@gmail.com wrote:
On Sun, Oct 6, 2019, 15:07 Miro Hrončok mhroncok@redhat.com wrote:
Couple of my updates have e-mailed me $subj. Is it something to worry about?
I got this too for a lot of my updates, just a few hours ago. I assumed it was caused by some kind of server glitch, maybe related to the current koji outage.
Got one as well, then got the following before I even started investi-gating:
This update's test gating status has been changed to 'ignored'.
Dridi
On Sun, Oct 06, 2019 at 03:11:28PM +0200, Fabio Valentini wrote:
On Sun, Oct 6, 2019, 15:07 Miro Hrončok <[1]mhroncok@redhat.com> wrote:
Couple of my updates have e-mailed me $subj. Is it something to worry about?I got this too for a lot of my updates, just a few hours ago. I assumed it was caused by some kind of server glitch, maybe related to the current koji outage.
It likely was. Bodhi called greenwave to ask about a status of the builds in an update, greenwave calls Koji for more information, that calls returned an error. Greenwave's request returns an error as well. Bodhi is not able to distinguish where the error is coming from and thus set the status to "greenwave_failed".
Pierre
On Mon, Oct 7, 2019 at 9:44 AM Pierre-Yves Chibon pingou@pingoured.fr wrote:
It likely was. Bodhi called greenwave to ask about a status of the builds in an update, greenwave calls Koji for more information, that calls returned an error. Greenwave's request returns an error as well. Bodhi is not able to distinguish where the error is coming from and thus set the status to "greenwave_failed".
So, this mean Bodhi will reject to move the package to updates or not? :-?
It's happening to me too[1].
[1] https://bodhi.fedoraproject.org/updates/FEDORA-2019-d9a4a2d770
On Tue, Dec 24, 2019 at 12:45:25PM +0100, Ismael Olea wrote:
On Mon, Oct 7, 2019 at 9:44 AM Pierre-Yves Chibon <[1]pingou@pingoured.fr> wrote:
It likely was. Bodhi called greenwave to ask about a status of the builds in an update, greenwave calls Koji for more information, that calls returned an error. Greenwave's request returns an error as well. Bodhi is not able to distinguish where the error is coming from and thus set the status to "greenwave_failed".So, this mean Bodhi will reject to move the package to updates or not? :-? It's happening to me too[1]. [1] [2]https://bodhi.fedoraproject.org/updates/FEDORA-2019-d9a4a2d770
Sorry for the late reply. For historical reasons, bodhi treats the "greenwave_failed" status as "passed", so it will not gate updates if greenwave failed to give it a go/nogo answer.
On the top of that, bodhi checks all un-pushed update when it calls greenwave which is why we sometime see this toggling. I have argued in the past that once an update passed greenwave we shouldn't ask greenwave if it changed its mind but the idea was then turned down.
Pierre
On Fri, 2020-01-10 at 14:37 +0100, Pierre-Yves Chibon wrote:
For historical reasons, bodhi treats the "greenwave_failed" status as "passed", so it will not gate updates if greenwave failed to give it a go/nogo answer.
It's not historical, or we wouldn't be discussing this - the message means that Greenwave returned a 5xx status code to Bodhi. If Bodhi didn't consider that a "pass", then we would be blocking updates due to systemic errors, and that would be unpleasant for packagers.
On the top of that, bodhi checks all un-pushed update when it calls greenwave which is why we sometime see this toggling. I have argued in the past that once an update passed greenwave we shouldn't ask greenwave if it changed its mind but the idea was then turned down.
It's not continuous integration if you don't do it continuously.
On Mon, Jan 13, 2020 at 09:30:33AM -0500, Randy Barlow wrote:
On Fri, 2020-01-10 at 14:37 +0100, Pierre-Yves Chibon wrote:
For historical reasons, bodhi treats the "greenwave_failed" status as "passed", so it will not gate updates if greenwave failed to give it a go/nogo answer.
It's not historical, or we wouldn't be discussing this - the message means that Greenwave returned a 5xx status code to Bodhi. If Bodhi didn't consider that a "pass", then we would be blocking updates due to systemic errors, and that would be unpleasant for packagers.
There are no more systemic errors, there are infrastructure issues at times which is what happened during the shutdown as we saw with other components.
On the top of that, bodhi checks all un-pushed update when it calls greenwave which is why we sometime see this toggling. I have argued in the past that once an update passed greenwave we shouldn't ask greenwave if it changed its mind but the idea was then turned down.
It's not continuous integration if you don't do it continuously.
We have the answer from the CI SIG saying that if the tests pass, we can assume they'll keep passing and optimize our script with this assumption.
Pierre
-
Dridi Boukelmoune -
Fabio Valentini -
Ismael Olea -
Miro Hrončok -
Pierre-Yves Chibon -
Randy Barlow