Comment 7 for bug 1401948

Answers:
1. The issue happened at scale lab with 3 controllers and 97 computes. Deployment HA + Ubuntu, without services (Ceilo, Sahara, Murano)
2. The issue happens once.

Since the issue is most probably occurs rarely and randomly, it would be sufficient to get the following info:
1. What do all these errors in rabbitmq log mean? is it possible to find from them what was wrong?
2. What kind of info we need to collect to simplify debugging in the future?

Take into account that the same may happen at customer deployment and they could certainly ask for reasons.