Proxied routes time out after 10 seconds
While investigating gitlab-org/gitlab-ce#33218, we noticed that Workhorse is timing out after 10 seconds with the message, "badgateway: failed after 10s: context canceled". Should this timeout be the same value as the unicorn timeout?
https://log.gitlap.com/goto/0a304c97561fad711c41d1cf1425c0e8
I don't see this in earlier Elasticsearch logs. Is this the result of moving to Go 1.8?
Edited by Jacob Vosmaer