[08:41:33] <hashar>	 Good morning, I could use a puppet-merge for https://gerrit.wikimedia.org/r/c/operations/puppet/+/758514 please.  That is for the WMCS CI instances.   
[08:41:35] <hashar>	 Yesterday E.mperor mentioned the patch could have used a +1 from another for releng and Ahmon did review the last iteration, so it is ready now :)
[08:43:23] <_joe_>	 hashar: it's a 200 loc patch, I guess anyone merging it will want to review it at least quickly, it will take some time
[08:46:19] <hashar>	 yeah I got it :]
[08:46:33] <hashar>	 though Ahmon and I paired on it extensively and it is already deployed :]
[08:48:54] <hashar>	 anyway there is no rush
[09:36:30] <RhinosF1>	 Hey, do we know of any throttling of the wikidata api? We got a user reporting slow response from a server they use to do a lot of queries but the actual page responds fine
[09:36:54] <RhinosF1>	 They're hitting esams
[09:38:09] <_joe_>	 no
[09:39:53] <RhinosF1>	 _joe_: do you know of any reason it would be randomly slow from one device
[10:07:36] <vgutierrez>	 RhinosF1: so... it they are hitting cp3062 they could be affected by https://phabricator.wikimedia.org/T300366
[10:07:42] <vgutierrez>	 *if
[10:08:16] <RhinosF1>	 vgutierrez: would that affect the api though
[10:08:24] <RhinosF1>	 I thought it was just thumbnails
[10:10:12] <vgutierrez>	 RhinosF1: per lego.ktm's friend comment mentioned in  https://phabricator.wikimedia.org/T300366#7677828 I'd say so
[10:10:42] <RhinosF1>	 vgutierrez: thanks
[10:35:25] <RhinosF1>	 vgutierrez: am I right that only 1 server runs envoy
[10:35:41] <vgutierrez>	 one server per DC and cluster
[10:36:07] <RhinosF1>	 vgutierrez: ty
[10:36:32] <RhinosF1>	 apparently some thread on commons (now archived) claims we dropped support for HTTP 1.0 entirely
[10:36:59] <vgutierrez>	 that's not accurate
[10:37:06] * RhinosF1 guessed
[10:37:09] <vgutierrez>	 for your reference, you can find the hostnames on https://phabricator.wikimedia.org/T271421
[10:38:26] <RhinosF1>	 ty
[10:44:28] <hashar>	 jbond: thx for the review :]
[10:56:25] <jbond>	 hashar: np
[11:05:51] <RhinosF1>	 vgutierrez: yep, moving to curl fixed it
[11:06:25] <vgutierrez>	 RhinosF1: we are in the process of replacing the envoy nodes with HAProxy ones later this week, so that should help
[11:06:56] <RhinosF1>	 ah ok
[12:40:50] <hashar>	 I am switching the operations-puppet-tests-buster-docker Jenkins job to a new instance (Stretch > Bullseye). Should be anoop
[13:05:37] <hashar>	 jbond: that checksum never being validated was a great catch :-]