Quantum agents fail at startup
When starting my one-node OpenStack machine, it appears that the quantum agent services stall and timeout when trying to access the rabbitmq server. This causes the DHCP and Level 3 network functions to fail. All I need to do is wait and restart the services, but shouldn't this happen after a specified time anyway? (I have waited 30-45 minutes, and it doesn't) Anybody out there with a similar issue that has a resolution? I am planning a lights-out datacenter and this kind of thing is a real issue.
Thanks, OldParrothead
Hi, what OS+level? Quantum package levels? Is it running on a VM and if so which? Which quantum plugin?
Thanks for the response. Running Ubuntu 12.04LTS NOT on a VM. Quantum plugin is openvswitch. Not sure of quantum versions, but both machines were downloaded and built since June 10th.
Sorry, I haven't seen that particular problem - I don't have an all-in-one. The cause maybe that the quantum agents and service are upstart jobs and rabbitmq is not. So they are starting before rabbit.
Thanks. Anybody else out there ? This is becoming a real stickler for me.
Check the hardware clock is not too far out - maybe a large adjustment by ntp at boot affects the agent retry code.