[11:03:15] <elukey>	 Hi folks, Wikifeeds seems to have a constant rate of error/50x since the last hours
[11:03:19] <elukey>	 https://grafana.wikimedia.org/d/lxZAdAdMk/wikifeeds?orgId=1&from=now-7d&to=now
[11:03:48] <elukey>	 it matches with an increase in traffic, so far the only thing that I can see is that the tls proxy is heavily throttled, but it didn't start today
[11:04:21] <elukey>	 here the app's logs https://logstash.wikimedia.org/goto/a0cbd0213ca07eb99a85f37f5f511c8d
[11:06:27] <elukey>	 to clear out any possible weird state of the envoy proxy, I am doing a roll restart of the pods
[11:13:03] <elukey>	 it seems going down, I'll check in a bit again.. maybe this is something similar to what we are seeing sporadically with Eventgate?
[11:16:21] <elukey>	 yep looks like it
[11:16:41] <elukey>	 in any case, the tls-proxy cpu throttling seems excessive: https://grafana.wikimedia.org/d/lxZAdAdMk/wikifeeds?orgId=1&from=now-90d&to=now&viewPanel=102
[11:17:04] <elukey>	 (no recent perf regression though, at least that I can see)
[11:17:34] <elukey>	 anyway, going afk again :)