[20:53:05] bd808: Any idea what's going on with https://phabricator.wikimedia.org/T111267 ? It looks like some Python script is running update.php many times in a loop and then barfing for no reason? [20:54:10] * bd808 reads the ticket [20:54:26] oh I hate that job [20:54:48] It hasn't worked in a long time apparently [20:55:16] update.php is failing... but why [20:56:15] Is it though? [20:56:15] I don't see it printing an error [20:56:28] I looked into this because of https://gerrit.wikimedia.org/r/235860 [20:56:42] But I didn't see the error message I would have expected in the Jenkins log; in fact I see no errors there at all [20:57:09] and on my machine update.php only got just past the countdown before barfing on a bad path there [20:57:48] hmmm.. I was seeing this `Exception: ('command: ', '/usr/local/bin/mwscript update.php --wiki=enwiki --quick', 'output: ', '')` and assuming that meant python got a non-zero exit code [20:58:25] maybe it's just timing out? /me goes to read the python driver script [20:59:03] Yeah I was also assuming a non-zero exit code [20:59:13] But then I got confused because update.php didn't appear to have screamed about anything [20:59:25] Then I got more confused when I found logs of like 3 or 4 consecutive update.php runs [21:07:51] That log message at the end does mean that the exit code of "/usr/local/bin/mwscript update.php --wiki=%s --quick" is > 0 [21:19:10] RoanKattouw: "Argument 7 passed to Flow\Model\WikiReference::__construct() must be an instance of Title, null given" [21:19:18] that's the crash [21:28:16] Ugh [21:28:18] THAT bug [21:28:27] Could you comment that on the task? [21:28:44] Actually I'll do it [21:28:49] I'm the one that asked for help in the first place [21:29:15] Oh wait you've already done it, even better [23:42:12] bd808: merged! [23:42:25] \o/ [23:42:49] one more bug and we'll be ready for a new release [23:43:35] is that https://github.com/wikimedia/composer-merge-plugin/issues/62 or mwjames's back-compat issue? [23:44:10] that #62 one [23:44:24] mwjames' issue is a non-issue IMO [23:46:44] ok, set the milestone for that issue [23:50:31] I'll look at it tomorrow/this weekend. There is a proposed patch already and I can reproduce the bug locally