[14:01:53] <_joe_>	 hi :) do you have a timeline on https://phabricator.wikimedia.org/T130692 ?
[14:02:22] <jynus>	 "jcrespo moved this task from Backlog to Next on the DBA board."
[14:02:41] <_joe_>	 yeah that I can see :P
[14:02:49] <jynus>	 So?
[14:02:55] <_joe_>	 I was wondering if you have an idea how long it would take
[14:03:19] <_joe_>	 because I have to time my work for https://phabricator.wikimedia.org/T86096 based on that :)
[14:03:47] <_joe_>	 tl;dr I have to rebuild hhvm for trusty with the newer libicu and when I do that it's an irreversible change
[14:04:09] <jynus>	 do you want me to delay or speed that?
[14:04:28] <_joe_>	 just know when do you think you might be done
[14:04:46] <_joe_>	 approximatively, like this week, next week, in a month
[14:05:00] <jynus>	 last week is the last estimation
[14:05:43] <jynus>	 that still stands
[14:10:47] <jynus>	 it says it is blocked on that, but that is false
[14:11:13] <jynus>	 even with that applied, appling that change to all wikis would fail without a script update
[14:11:30] <_joe_>	 uhm
[14:11:47] <_joe_>	 ok that was not clear at all :)
[14:12:18] <jynus>	 and even without it, if the update script was updated, it would not be an blocker
[14:13:05] <jynus>	 my change only makes it easier
[17:22:22] <Glaisher>	 jynus: Re T134136 what was the old name/new name there?
[17:22:22] <stashbot>	 T134136: Heavy query from CheckUser extension caused replica lag on all S1 slaves - https://phabricator.wikimedia.org/T134136
[17:22:33] <Glaisher>	 Renames are publicly logged so it's not private data
[17:23:12] <jynus>	 not sure if volans_ will have that info more handy
[17:23:27] <jynus>	 if not, I can try to get it again from logs
[17:26:03] <jynus>	 ah
[17:26:14] <jynus>	 I think the 0 was sanitized data
[17:26:46] <jynus>	 so it was my wrong interpretation
[17:27:27] <jynus>	 or not?
[17:27:51] <jynus>	 no no, it is 0
[17:28:02] <volans_>	 Glaisher: yes I have it
[17:28:34] <jynus>	 volans_, it is this one, right? https://tendril.wikimedia.org/report/slow_queries_checksum?checksum=4ae85c505b6b0e6fe4fe7bf7f8f57c98&host=db1057&user=wikiuser&schema=enwiki&hours=48
[17:28:47] <jynus>	 if yes, I can continue, go back to your business
[17:29:25] <volans_>	 jynus, Glaisher yes confirmed
[17:29:38] <volans_>	 is that one
[17:30:09] <jynus>	 so, I can share the full query to NDA-signed people if you are interested
[17:30:48] <jynus>	 it actually happened twice
[17:31:09] <Glaisher>	 Does the other RenameuserSQL queries for that user have the same 0 id?
[17:31:20] <jynus>	 yes
[17:31:28] <jynus>	 although I think it is not an id
[17:31:50] <jynus>	 cuc_user is an foreign key or a flag?
[17:32:21] <Glaisher>	 It's a foreign key for user.user_id but there is no actual constraint on the db side
[17:32:29] <jynus>	 of course
[17:32:33] <jynus>	 I meant conceptually
[17:33:19] <jynus>	 at this point I would turn that into a security issue/UBN, but your call
[17:33:48] <jynus>	 let me paste the too instances on a resticted paste
[17:33:59] <Glaisher>	 I think legoktm is in a better position to decide on that :)
[17:34:11] <legoktm>	 hi
[17:34:14] <Glaisher>	 (cc'd him)
[17:34:16] <jynus>	 as in any case it contains private info
[17:34:18] <legoktm>	 link?
[17:34:26] <Glaisher>	 https://phabricator.wikimedia.org/T134136
[17:36:42] <jynus>	 I've shared the instances I found, I think now it is in your hands
[17:37:43] <legoktm>	 jynus, Glaisher: I marked the task as security and commented
[17:38:33] <jynus>	 legoktm, thanks
[18:57:52] <jynus>	 db1040 new jessue/mariadb 10 host