[07:27:10] hello hello [07:27:16] pcc runs end up in puppet_compiler.controller.ControllerNoHostsError [07:27:19] is it known? [07:27:47] example https://integration.wikimedia.org/ci/job/operations-puppet-catalog-compiler-puppet7-test/lastBuild/console [07:36:00] ah wait this may be Hosts: auto [07:36:15] uff nevermind, it doesn't find the targets, okok [08:03:10] FYI, I'm disabling Puppet merges for approx 15 mins for Puppet server reboots [08:19:17] Puppet merges are re-enabled [09:56:41] I broke the tcp_kafka_broker_tls_ip4 check, revert on the way [23:00:44] @rzl you around? [23:00:51] 👋 [23:01:38] There is a thread in #talk-to-sre on slack. And I would love a quick triage from an SRE just in case there is a potentially bigger underlying issues. [23:06:53] I was looking at that briefly earlier, I don't know much but I can drop a comment on the thread! no wrong answer, just curious -- is there something about that question that's particularly rzl-oriented out of all the SREs, or am I just a friendly face? :) [23:08:51] rzl: A friendly face and I didn't know who else might be awake given we are closing in on end of day for the remainder of the west coast folks you were also one of the first people I thought of that might have been around still. [23:09:55] ah cool, thanks :) still no wrong answer [23:10:19] the spoiler is I don't think I'm going to have anything decisive, and we should dig a little further in a "High but not UBN" kind of way [23:11:27] (for anyone not on slack, the gist is, "hey the app layer sure is getting a lot more traffic since about 02:00-04:00 on 2026-09-01, and that sure is impacting PHP-FPM saturation and response latency, what's up with that" and so far afaik the answer is just "yeah it sure is, what's up with that") [23:12:00] cc jasmine_ and cwhite as oncallers ^ [23:14:16] thanks rzl & seddon [23:15:40] yeah was looking at this earlier too, very brief discussion in #mediawiki_security, mentioning MySQL query throughputdoubled on s1 and ecordLintjob and parsoidCachePrewarm trailing, but taking a deeper look [23:18:07] that was an unrelated template change I think, this is something that started earlier [23:22:45] oh right that make sense, seeing that the template-radar error note now too