Comment 2 for bug 1331274

Revision history for this message
Sean Dague (sdague) wrote : Re: Couldn't find log for * at /opt/stack/new/screen-logs/*

I don't think this is a new bug, this is a move of an existing set of issues with extremely slow nodes. Realistically once you filter out the neutron parts of this failure you get only a handful of hits. (6 jobs in the last 24 hrs)

query: >
  message:"Expected running services not running:"
  AND NOT message:"q-fwaas"
  AND tags:"console.html"

In all but 1 case they are hpcloud. They are split across grenade and grenade-partial pretty evenly (I had wondered if partial was more susceptible, it's not.)

In most of these cases it seems like these services are actually running, but logs don't exist. This actually exposes another long term mystery we were seeing where logs were not created. So I think we need to move forward here to figure out why and get to the root cause.