[15:30:46] <Krinkle>	 RainbowSprinkles: bd808: What's the current status of https://phabricator.wikimedia.org/T89169 and https://phabricator.wikimedia.org/T45086 ? I thought they were resolved now, right? At least the fatals have stacks, but also the warnings. Or did it break again / not "completely" finished? Unclear to me what's left to do.
[15:30:57] * Krinkle is triaging https://phabricator.wikimedia.org/maniphest/query/Vc.M7oRZYxUR/#R
[16:23:40] <bd808>	 Krinkle: I did all I could do with those quite a while ago. If things are still not what people want/need someone else needs to pick it up and try to find a solution
[16:23:57] <Krinkle>	 bd808: Aye, yeah, no problem. Would like to help actually.
[16:24:04] <Krinkle>	 Just somewhat unsure what the situation is at the moment.
[16:24:30] <bd808>	 *nod* I'm really not sure either
[16:24:31] <Krinkle>	 Afaik fatals always had traces, except when they didn't for a shor ttime, and that was restoed. Warnings never did in prod until we started logging the error channel from within mw
[16:24:47] <bd808>	 right
[16:24:57] <bd808>	 if hhvm logs it we won't get a trace
[16:25:13] <bd808>	 the only way to fix that is custom hhvm c++ code
[16:28:56] <Krinkle>	 bd808: Right, the hhvm log is mostly redundant just for cases where mw might not be able to log it
[16:29:11] <Krinkle>	 which seems unlikely as that should be a fatal error, which we have traces for.
[19:03:22] <greg-g>	 Krinkle: this was a fatal that we didn't get a backtrace for (until someone reproduced it via the api): https://phabricator.wikimedia.org/T168681
[21:10:42] <hasharAway>	 the task being https://phabricator.wikimedia.org/T89169  though we have some in fatal.log