[08:49:58] * kostajh waves [08:50:19] would someone be able to have a look at T306349#8382132 and comment on there from SRE's perspective? [08:50:20] T306349: Public-facing API for image suggestions data - https://phabricator.wikimedia.org/T306349 [08:50:40] Oops, I just realized I never added a SRE phab project tag, so I'll do that now [08:56:26] <_joe_> kostajh: sure [09:01:17] heads up -- I'm starting the graphite/statsd failover to graphite1005 [09:02:51] <_joe_> godog: ack [09:03:54] <_joe_> kostajh: I'd remove " Additionally, a public-facing API will be helpful with QA and testing." first of all :P [09:04:13] <_joe_> given the public facing API in this case will be production :) [09:06:20] _joe_: I revised it, what it's trying to express is: for MediaWiki extension code, currently we either use mock data, or use SSH tunnels through mwmaint to work with the image suggestion API, but a public proxy would allow us to talk to the service [09:51:11] Can I get a review of https://gerrit.wikimedia.org/r/c/operations/puppet/+/862175 ? [09:52:09] marostegui: on it [09:52:22] volans: :* [09:53:57] {done} [09:54:30] thank you! [09:57:15] anytime ;) [10:24:52] <_joe_> kostajh: yeah the problem is that a public proxy should either be application-aware and avoid exposing any write endpoints [10:25:06] <_joe_> or use some for of auth to allow you to access it [10:25:28] <_joe_> and I don't think the api gateway is the right place where to create a debugging proxy for developers [15:52:47] elukey: klausman: papaul wants to power off ores2009 for https://phabricator.wikimedia.org/T323925 [15:53:32] akosiaris: ack doing it [18:02:44] has anyone encountered this error when using docker-pkg? [18:02:44] 2022-11-30 12:57:17,273 [docker-pkg-build] INFO - W: GPG error: http://security.debian.org buster/updates InRelease: At least one invalid signature was encountered. [18:02:44] E: The repository 'http://security.debian.org buster/updates InRelease' is not signed. [18:02:55] happens during an attetmpted apt-get update [18:03:06] been googling and trying various docker prune commands, but not sure what is going on [18:06:34] ottomata: I think I did see that. docker prune is the right lines, but I think I had to remove the complete base image with `docker rmi` [18:20:35] I'm trying to find it in the IRC logs but can't. [18:32:26] oh i'm just buidling locally atm [18:32:39] btullis: b ase image being which one? [18:39:05] i did a full docker system prune -af --volumes, which causexd the debian images, to be redownloaded [18:39:13] but still same error [18:39:19] is it possible some clock is wrong on my docker host? [18:41:04] this worked for me 8 days ago, but it looks like the wikimeda debian images were updated 3 days ago? [18:41:43] clock looks fine on my host [19:11:23] ok turns hout 50G was just not enough for the docker disk size for bulding images :) Bumped to 100G and things are working better