kocolosk commented on issue #548: [Jenkins] 500 error on getting view output
URL: https://github.com/apache/couchdb/issues/548#issuecomment-313228860
Not a blocker. We should write a solid reproducible test case for it (should
be easy enough in EUnit right?) and address it, but agree it's d
kocolosk commented on issue #548: [Jenkins] 500 error on getting view output
URL: https://github.com/apache/couchdb/issues/548#issuecomment-310541660
The more I dig the more I think this is a non-trivial race condition. I'd
recommend that we file a separate ticket explicitly focused on addr
kocolosk commented on issue #548: [Jenkins] 500 error on getting view output
URL: https://github.com/apache/couchdb/issues/548#issuecomment-310508601
Hmm, I didn't need to do anything special -- just ran it in a bash loop and
it would fail within the first 10 runs.
---
kocolosk commented on issue #548: [Jenkins] 500 error on getting view output
URL: https://github.com/apache/couchdb/issues/548#issuecomment-310504265
@eiri I'm curious about your thoughts on this. You were poking at some of
the same issues in #576 but I think that didn't totally address the
kocolosk commented on issue #548: [Jenkins] 500 error on getting view output
URL: https://github.com/apache/couchdb/issues/548#issuecomment-310491549
I think I know what's happening here. Every `db.query()` invocation in this
test creates a ddoc with a random name, queries the view, and the
kocolosk commented on issue #548: [Jenkins] 500 error on getting view output
URL: https://github.com/apache/couchdb/issues/548#issuecomment-310460143
I thought that perhaps with the work in #576 and #577 this might be
resolved, but now I think we may have just morphed to an `{'EXIT',normal}
