[08:33:20] <zpapierski>	 off to bank, be back around noon
[08:54:19] <ejoseph>	 dcausse: someone suggested to use this in place of vagrant, https://gitlab.wikimedia.org/mhurd/mediawiki-cirrus-docker
[11:04:10] <ejoseph>	 break
[12:08:27] <zpapierski>	 mpham: usage data for beta 1 and 2 uploaded for the last 14 days
[14:06:29] <inflatador>	 greetings
[14:06:33] <zpapierski>	 o/
[14:23:45] <inflatador>	 ejoseph just got your invite, I will be there!
[14:42:31] <ejoseph>	 Thanks
[15:53:52] <ebernhardson>	 \o
[16:02:46] <gehel>	 ejoseph: triage: https://meet.google.com/qho-jyqp-qos
[16:03:07] <ejoseph>	 joining
[18:09:39] <inflatador>	 gehel or anyone else, I finished https://phabricator.wikimedia.org/T301954 , let me know if 'resolved' is the proper status or I need change anything else on it
[18:10:54] <inflatador>	 actually, check that, I had to reopen as it's not truly complete until puppet changes are finished
[18:46:29] <inflatador>	 lunch/errands, back in ~1h
[19:15:06] <ebernhardson>	  thats cool, pickaxe (-S) works on git reflog  too
[19:18:02] * ebernhardson apparently tried to lose a commit but it was just ~80 lines back in reflog
[19:27:30] <gehel>	 ryankemper: I'll be 5' late
[19:28:03] <ryankemper>	 gehel: ack
[19:30:50] <inflatador>	 aaand back
[19:31:45] <gehel>	 ryankemper: I was wrong. Augustin was more collaborative than expected. I'm already there!
[21:19:49] <ebernhardson>	 huh, i wouldn't have guessed our internal domain names are visible from reverse dns: https://network-tools.webwiz.net/reverse-dns.htm
[21:20:14] <ebernhardson>	 err, i guess that link doesn't include options :P but i put in `2620:0:860:104:10:192:48:45` which finds wcqs2003.codfw.wmnet
[21:20:34] * ebernhardson was at first surprised to see my own dns return it, and had to try somewhere else to be sure i don't have some wierd local config P
[21:32:33] <ebernhardson>	 was able to trigger an auth problem on commons-query, might be the same or at least related to the issue we are already trying to solve. The problem, at least in this repro, is that the request is blocked by CORS policy: Response to preflight request doesn't pass access control check: Redirect is not allowed for a preflight request
[21:32:44] <ebernhardson>	 not sure how to fix, i guess next step is read the mdn docs on CORS :P
[21:33:56] <ebernhardson>	 i don't think so though, the underlying idea here is to prevent one site from pretending to be the other one unless the other explicitly allows it
[21:36:17] <volans>	 ebernhardson: that's because of IPv6, that's a "public" ip ;)
[21:36:24] <volans>	 (the resolution)
[21:38:19] <ebernhardson>	 volans: interesting, i guess i "know" there are that many ipv6 addresses but don't usually think about it :)
[21:54:11] <inflatador>	 and back
[21:54:36] <inflatador>	 IPv6 and RDNS can be fun ;P
[22:56:15] <ebernhardson>	 There is some "law" of computing that basically says if you fan out a request to 100 shards and wait for them all to complete, You are basically getting the 99th percentile latency of the underlying requests as the mean of the top level request. Does that have a particular name?
[23:41:03] <ryankemper>	 ebernhardson: good question :P no idea, but if there's not one you should definitely claim that law
[23:41:30] <ryankemper>	 the only downside is all the credit will probably go to that other erikbern.com guy xD
[23:44:52] <ebernhardson>	 lol, i'm sure i've read this before but maybe it was just some documentation about go-routines and fanning out requests or some such
[23:53:00] <inflatador>	 ryankemper as promised, here's the MR. We'll see if CI approves https://gerrit.wikimedia.org/r/c/operations/puppet/+/766876
[23:53:52] <inflatador>	 busted on the line length...trying again