Hello!
Yesterday around 21:43 UTC we experienced a slowdown of Elasticsearch.
In short, one of our elasticsearch server went crazy with its garbage
collector and had to be restarted.
You can find more details in the incident report [1] or the
corresponding Phabricator task [2].
Thank you all for your understanding!
Guillaume
[1]
https://wikitech.wikimedia.org/wiki/Incident_documentation/20160509-CirrusS…
[2]
https://phabricator.wikimedia.org/T134829
--
Guillaume Lederrey
Operations Engineer, Discovery
Wikimedia Foundation