Ask Your Question

NicCOConnor's profile - activity

overview network karma followed questions activity
2013-09-02 07:33:10 -0600 received badge ● Good Question (source)
2013-08-01 11:18:21 -0600 received badge ● Famous Question (source)
2013-08-01 11:18:21 -0600 received badge ● Notable Question (source)
2013-08-01 11:18:21 -0600 received badge ● Popular Question (source)
2013-07-15 15:53:26 -0600 received badge ● Nice Question (source)
2013-06-19 15:40:20 -0600 received badge ● Famous Question (source)
2013-05-28 00:27:07 -0600 received badge ● Notable Question (source)
2013-05-26 12:39:08 -0600 received badge ● Popular Question (source)
2013-05-21 13:59:41 -0600 received badge ● Student (source)
2013-05-21 13:30:35 -0600 asked a question mysqld server 100% cpu

So I made a bit of an error while testing out my system. I have a pretty powerful blade running all services, and I wanted to stress test it by rolling out 80 instances. this seems to have caused mysqld to crap out and utilize 100% of the processors which is pretty impressive since there are 4 XEON 10 core processors in the blade. I believe this could have something to do with quantum and it's default quota of 50 ports, however I've been at it all day and I can't seem to figure out how to reset the system. I've tried deleting all the instances and the process is slow and error prone. Almost every VM is in an error state or deleting state.

nova list takes about 20 minutes to run. horizon takes about an 20 minutes to list instances almost all are in error or deleting state. mysqld utilizing 100% cpu (htop) tried clearing rabbitmq-queues for nova and quantum. no luck. Reboot server has same issue. I've checked the logs but I really can't decipher what exactly is holding up the system. The system is offline so I can't really post logs, without jumping through some bureaucratic hoops.

Right now I'm contemplating dropping the nova and quantum databases and re-initializing them but I'm not sure what the adverse effects are.

Can anyone shed some light on how to resolve an issue like this.

Nic

2013-05-16 14:10:33 -0600 answered a question novnc - Horizon iFrame refuses to update novnc address

Did a diff of a known good config nova.conf had the wrong setting. vnc_enable=true was inputted incorrectly novnc_enable=true.
2013-05-16 09:34:02 -0600 received badge ● Editor (source)
2013-05-16 09:18:35 -0600 asked a question novnc - Horizon iFrame refuses to update novnc address

I have a single machine running an entire Grizzly stack with Keystone, Glance, Cinder, Quantum, Nova, Nova-compute and Horizon.

I can launch instances and everything seems to be working as it should however I cannot get horizon dashboard to update my novnc address. It keeps defaulting to 127.0.0.1

I have made the settings changes in nova.conf, nova-compute.conf, api-paste.ini

If I click the link "show only console" I get the same issue, however I can change 127.0.0.1 to the correct IP and I get the console and everything works as it should so my authentication, tokens and process seems to be working as they should it's just the horizon IFrame that refuses to update the novnc IP

I've tried search and grepping for any instance of novnc or 127.0.0.1 but I just can't find it anywhere, I found the launchpad page at https://bugs.launchpad.net/horizon/ listed under openstack-dashboard. I did not see any bugs related to novnc. versions for novncproxy and openstack-dashboard(horizon) are both 2013.1

UPDATE: I tried to uninstall openstack-dashboard apt-get remove --purge openstack-dashboard. I reinstalled ensuring all settings were correct in the conf files with no success.

Nic