[13:12:15] <arnaudb>	 hey, I'll be afk this afternoon I'm not feeling too well
[13:12:47] <Emperor>	 :( GWS!
[16:21:07] <marostegui>	 I can't believe this bug isn't fixed yet
[16:21:08] <marostegui>	 https://jira.mariadb.org/browse/MDEV-12572
[16:21:12] <marostegui>	 I am going to comment on it
[16:23:33] <jynus>	 is it just the syntax or is the algorithm not supported?
[16:24:05] <marostegui>	 the syntax, cause removing an index in place works fine with alter table
[16:24:34] <jynus>	 interesting because mariadb was all about unifying the syntax a few versions ago
[16:24:56] <jynus>	 (the IF EXISTS, etc.)
[16:26:47] <marostegui>	 yeah, I was issuing a drop index if exists and I started to see the table locked and I was like :wtf?
[16:26:50] <marostegui>	 And found that
[17:14:05] <Amir1>	 https://www.irccloud.com/pastebin/yUBtwBzn/
[17:14:21] <Amir1>	 we probably can run the img_id schema change with replication everywhere except s4
[17:15:22] <Amir1>	 (we can also set the whole cluster RO too)
[17:19:29] <marostegui>	 yeah, those sizes look relatively okay
[17:43:44] <marostegui>	 We really need to get better at keep tracking of all the depooled hosts. 
[17:45:36] <marostegui>	 There are hosts there which have been depooled like 20 days ago
[17:45:44] <marostegui>	 And never got repooled
[17:45:45] <marostegui>	 :(
[17:46:29] <marostegui>	 I am going to start repooling them, I see no reason why they are depooled
[17:47:50] <volans>	 you know, you could add a check :-P
[17:49:17] * volans hides as this was a low one