[06:51:37] hello folks [06:51:55] bfd has been flapping between drmrs and eqiad (GTT transport afaics) [06:52:52] elukey: yeah, I'm in the process of opening a Juniper ticket [06:52:58] something weird is going on [06:53:28] XioNoX: ah nice ack :) [06:53:45] thanks for keeping an eye though [06:59:58] your fault, you taught me how to check weird alerts on routers :D [07:10:43] elukey: I think I fixed it, writing down the JTAC ticket gave me ideas [07:13:12] or maybe just one of the issues [07:35:02] elukey: I opened https://phabricator.wikimedia.org/T321034 (cc topranks) [07:37:27] Huh [07:37:34] Odd issue [07:38:42] yeah, I was happy to fix it at first, and then the v4 issue... [07:40:32] weird [08:46:12] XioNoX: I seem to have got that BFD to stabalize [08:46:19] can't explain why though https://phabricator.wikimedia.org/T321034#8324005 [08:49:48] good old turn off/on again [08:49:50] nice [08:50:00] I've largely built my career on it :P [08:51:42] As you hinted at perhaps something was "stuck" in the internals tying the address/session to the old int. [09:06:52] everything is indeed stable now, I removed the Icinga downtimes [16:29:50] <_joe_> !issync [16:29:51] Syncing #wikimedia-sre (requested by joe_oblivian) [16:29:53] Set /cs flags #wikimedia-sre sirenbot -Aitv [16:58:58] _joe_: is see on the RED dash that PHP 8.1 is already installed on some servers, awesome :) [17:00:27] can you confirm 'officialy' why 8.1 is the next version? Afaik we don't use debian upstream package, so that's presumably not the reason. Was it informed by MW compat (e.g. PHP 8.2 might be too big a jump to maintain concurrent compat in CI with 7.4 for the same dependencies, PHP 8.0 might be possible but why two migrations if we can go direct etc.) or was it a different reason? [17:00:50] There's a question at https://phabricator.wikimedia.org/T319432 and I imagine we'll get the question more often, so looking to document it, not really questioning it now. [17:04:28] <_joe_> Krinkle: I'll want to discuss this with moritz next week, but I assumed 8.1 was the sweet spot of being in debian testing AND being what interested the developers [17:04:50] <_joe_> so it's going to be the "official version" somewhere down the road [17:05:09] <_joe_> (and yes, I hope with mw-on-k8s such transitions will be easier to manage [17:16:23] php.net says 8.2.0 is in RC4, with a note of "Please DO NOT use this version in production, it is an early test version." [17:16:52] https://wiki.php.net/todo/php82 says GA: Nov 24, 2022 [17:20:10] Debian has (from rmadison): [17:20:12] php8.1 | 8.1.7-1 | testing | source, all [17:20:14] php8.2 | 8.2.0~alpha3-1 | experimental | source, all [17:22:15] <_joe_> paravoid: we will definitely not move to php 8 before 8.2 is GA [17:22:18] <_joe_> but still [19:08:44] legoktm[m]:it worked.https://phabricator.wikimedia.org/T297791 is now public [20:02:35] mutante, inflatador: still a few minutes away from home but I'm back online and have my laptop on me, thanks again for the coverage [20:03:31] rzl: alright:) np [20:04:37] ACK [20:18:48] \o/ [23:01:51] rzl, inflatador: happy end-of-shift. since we are off now, I will keep driving and be generally unavailable. I started writing a report for yesterday but yet have to finish it. cheers [23:02:02] just where is the bot topic change, heh [23:02:08] hand-over for others: nothing new this shift [23:02:14] usually comes in at three or four minutes past the mark, nbd [23:02:22] sirenbot does, that is [23:02:23] ah, gotcha:) [23:02:35] or we can do this: [23:02:36] !refresh-topic [23:02:45] ahem. [23:02:51] well, or we can wait! :D [23:03:05] I had no idea it was already "chatops" [23:03:11] heh [23:04:17] \o/ [23:05:25] :) alright, cya [23:05:27] (I think it's just running in an "every five minutes" refresh loop rather than aligned to :00:00 exactly, so you get a few minutes delay depending on what time the bot happened to start up) [23:05:33] ah, that makes sense [23:05:34] have a good evening, enjoy the drive! [23:05:41] good evening likewise [23:23:35] {◕ ◡ ◕}