On Tue, 2017-04-18 at 10:18 -0600, Kevin Fenzi wrote:
On 04/18/2017 10:08 AM, Adam Williamson wrote:
> One thought: we actually have the same problem for openQA, now it's
> running tests on updates. This isn't very visible at present so devs
> haven't been asking about it, but I suspect as soon as the results show
> up in Bodhi, someone will be wanting a 're-run test' button. I'm not
> sure how far we could share any part of implementing it, though...
So what is currently triggering the tests now?
fedmsgs when builds are tagged into specific koji tags?
I know it's not a fancy gui, but could we just untag/retag something to
emit another fedmsg to retest?
Dunno about Taskotron package-level tests. openQA update-level tests
trigger on 'bodhi.update.edit' and 'bodhi.update.request.testing'
messages; you can trigger a re-run of *all* the tests by editing the
update, but that's a bit silly.
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net