[16:58:32] jynus: springle: ping [16:58:36] Urgentish [17:02:14] tell me, hoo [17:02:39] I suspect s7 to be corrupt (missing a row in centralauth.globaluser) [17:03:23] this is not a good channel, it is publicly logged [17:04:03] I checked all slaves, might be memcached acting up [17:04:22] well, I need more info [17:04:32] to help you [17:05:03] Let me try poking memcached [17:05:15] I couldn't identify any of the slaves [17:05:23] but it's flaking, so that's weird for mc [17:06:06] can you please privately tell me what row you are looking for? [17:06:13] Yeah [17:06:17] ah crap [17:06:22] was looking at the wrong table [17:06:26] m( [17:06:38] purging hte mc key didn't help [17:14:05] No idea for now [17:14:19] I'll poke further later on, sorry (can't nail it down to anything) [17:14:30] so, to make it clear [17:14:36] the problem exists [17:14:54] but you cannot say it is the database for sure, correct? [17:15:34] in any case, add me to a ticket and I will be glad to help [17:15:44] but I need to know more :-) [17:27:16] (user sent me more info by PM) [17:30:45] hoo, this is T102915, I will give it a look later [17:41:09] I looked into that a few months ago and never found anything out [17:41:49] yes, I am only "comvincing" hoo that it is not an unbreak now [17:42:07] because it is been there for some time and nothing horribly has happened [17:42:25] yeah, nothing was wrong outside of that special page afaik [17:42:28] I may give it a try tomorrow in any case [17:45:29] specially to discard data consistency problems, with would have a larger impact [17:53:53] I would take a poke at it on mw1017, but it looks like there's a deployment scheduled in a few minutes [17:54:42] var_dump( gethostbyaddr( GlobalUsersPager::mDb->getServer() ) ); etc. [17:54:50] except obviously in a way that would actually work [17:58:19] in any case, lets avoid discussing non-db things (even if this happens to be the database :-)) here and do it on the main channels