Hey Daniel, Are you using the Mesa 1.6.1 release? There were 2 networking issues that we found and corrected while testing Mesa 1.6.1. Not sure if these issues exist in Mesa 1.6 or not, but if you are using Mesa 1.6, then switching to Mesa 1.6.1 may resolve your problems. The issues involved problems that occurred when rebooting either the nova-compute nodes or the node running the quantum server.
Thanks, Chris Dell From: crowbar-bounces On Behalf Of Williams, J T Sent: Thursday, November 07, 2013 3:51 PM To: dsanpe...@mcna.net<mailto:dsanpe...@mcna.net>; crowbar Subject: Re: [Crowbar] crowbar dying and OpenStack QuantumClient Exception Can you provide the information from running the command "netstat -nr" and traceroute for the following scenarios. the controller node can no longer ping or talk to the Network node from controller node run traceroute <network node IP> network node can no longer ping anything on the public network from the network node run traceroute <any known working public network IP> From: crowbar-bounces On Behalf Of Daniel San Pedro Sent: Monday, October 28, 2013 11:09 AM To: crowbar Subject: [Crowbar] crowbar dying and OpenStack QuantumClient Exception Good morning! My current setup: * crowbar admin node: PowerEdge 1950 2GB RAM * Controller: PowerEdge 2950 2GB RAM * Network (Quantum) node: PowerEdge 1950 2GB RAM * Compute and Cinder: Dell R720 64GB RAM 4TB Storage * Upon initial deployment, everything works great! Problems: * After a certain amount of time (8-12) hours, crowbar server slows to a crawl on SSH with very high iowait times, and crowbar UI is unresponsive. Either a manual kill of chef-client revivies the server or a reboot (chef-client restart sometimes does not work). Other problem if no high iowait times is that one or several nodes no longer show green on UI. I have to SSH to each individual node and either restart chef-client or kill it and start it again, or reboot the crowbar machine. I guess after so many restarts crowbar no longer stays live for more than 5 minutes! * After a week (or perhaps a weekend) OpenStack utterly breaks with no intervention on my part. No high iowait times. First, apache stops. I have to manually kill keystone and then start apache, as apache complains that 0.0.0.0:5000 is already taken (by keystone). Once apache starts, I get the following error: ...
_______________________________________________ Crowbar mailing list Crowbar@dell.com https://lists.us.dell.com/mailman/listinfo/crowbar For more information: http://crowbar.github.com/
