On Wed, 2018-07-11 at 13:59 -0700, Adam Williamson wrote: > On Wed, 2018-07-11 at 19:03 +0000, rawhide@xxxxxxxxxxxxxxxxx wrote: > > According to the schedule [1], Fedora 27 Candidate Update-20180711.1812 is now > > available for testing. > > Uhhhhhhh...OK? > > This obviously shouldn't have happened. I'm not sure what this compose > is or why it somehow passed all the relvalconsumer heuristics to get a > validation event created for it, but I'll try and figure it out. Sorry > for any inconvenience. I'll reset all the 'current' things in the wiki > to point back to the last proper validation event, please ignore all > this "event's" pages The wiki should be back to normal now. For anyone who's interested, the story here was slightly interesting: it turns out that ever since releng started doing 'updates' and 'updates-testing' composes, relvalconsumer has only been prevented from creating validation events for these by a single version check, because wikitcms' function for parsing a compose ID and determining the corresponding validation event was parsing those composes incorrectly as being Branched candidate composes for the release in question (27 or 28). relvalconsumer is set to only create events for "the release after the current release" - so, 29 at present - and that's been preventing these events from actually getting created...but yesterday there was a blip and the data source fedfind uses to decide what the "current release" is was giving wrong information...it said the current release was 26 :) So relvalconsumer thought it was creating an event for a Branched candidate for the next release, and went ahead. The weird '1812' version component is the *label* of the updates- testing compose, for anyone who wondered :P For real candidate composes, the label is the '2' in Beta-1.2 or the '3' in Beta-1.3. In this case wikitcms decided we had a 'Update' candidate compose versioned '20180711.1812'...hehe. I'm sending out a fix for wikitcms to make it parse these composes correctly (or rather just reject them as we never want validation events for them) which should prevent anything like this happening again. -- Adam Williamson Fedora QA Community Monkey IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net http://www.happyassassin.net _______________________________________________ test mailing list -- test@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to test-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/test@xxxxxxxxxxxxxxxxxxxxxxx/message/RJNT7YB4VWULXDESC5WTQ6LF2PFMAJJ7/
