[00:32:58] do you think db1022 deserves a pt checksum or sync? we havn't recently checked dbstore1001, and it seems like a good opportunity for sanity check [00:33:13] jynus_: (just $0.02) [02:26:27] jynus_: nm, my mistake. i see what you did [05:17:49] db1034 is me. normal upgrade. possibly also reload and/or switch it to logpager [06:58:44] I will profile db1018 as I did last week for 24 hours to assess the impact of P_S [09:49:14] the lag on 1069:3 is cause by a pt-table-checksum - many tables do not exist there from testwiki [17:52:53] jynus: do you need tendril right now, maybe 2 minuts downtime [17:53:06] i want to make sure it gets cloned right from WMF rpo [17:53:26] by removing the existing files and letting puppet add them again [17:53:35] not a good time now [17:53:57] it is the only place where I can monitor access failures for non-production [17:54:14] can you wait some minutes, maybe 1 hour [17:54:25] ok, i won't touch anything else [17:54:39] i did merge a puppet change but that was no-op [17:54:48] of course, it can wait [18:23:14] mutante, I think I can survive now without it, if you promise to put it up again :-) [18:23:24] please make a backup [18:23:45] I am not 100% sure that the repo and the files there would be the same [18:24:07] (I tried to deploy it on my pc and I lacked of some files) [18:25:08] jynus: which repo did you clone from? github or gerrit? [18:25:17] github [18:25:36] so yea, i did make a local backup [18:25:51] but if github repo is not complete.. then sigh [18:25:58] i should delay this then [18:26:15] not sure if because of the files, though [18:26:24] it may be because the lack of sql scripts [18:26:45] hrmm.. i'll try it. and if it fails then rsync back from backup [18:26:50] I didn't answer the gerrit because I wanted to check for differences first [18:27:07] but too many pending things [18:33:06] trying it now [18:38:21] it seems back [18:38:34] jynus: restored from local backup [18:38:41] because.. i did not clone it .. at all [18:38:46] oh [18:38:55] even though we use the same for other tools [18:38:58] maybe I can help tomorrow [18:39:03] or to be fair [18:39:06] so something is broken with puppet git clone in general :/ [18:39:12] springle is your guy for this [18:39:15] or i made a mistake [18:39:27] yea, so.. i will hold back on this for now [18:39:33] and first test the cloning somehow [18:39:38] and probably continue tomorrow [18:40:01] sorry for this, that's not what i expected at all [18:40:14] maybe some permission issues or something.. but not nothing happening at all [18:41:00] do not worry, I do not know about sean, but I havn't touched that code yet [18:41:14] we plan to do it soon [18:41:31] alright, so it worked in the past for dbtree [18:41:34] we did the same thing [18:41:40] something must have changed in between [18:41:46] will figure it out, but not right now [18:41:59] so you can go ahead