04-28-2014 06:18 AM
I'm doing some fancy cloning,
we have several clusters, same HW, so I want to reuse the image
from one node to another cluster, same disklayout etc.
I do a dummy install with the right disk partitioning.
Then I replace the image files partachi-sda*.tar.bz2
Deploy image ok, and runs as expected, so far so good, this works fine on a OS level.
But CMU does not like this, the SMD does not start on the nodes.
I can get it to start if I do (in the CMU GUI) Monitoring -> stop then start Monitoring Engine.
But when I reboot a node SMD doesn't start on that node.
Any advice ?
04-29-2014 06:46 AM
How many nodes do you have in your cluster?
Are all nodes part of network entity ?
Do you mean that every time when you reboot a compute node, you need to restart monitoring to get SMD running on compute node?
did you wait for few minutes after node boots successfully?
Please Increase monitoring debug level CMU_SEC_MONITORING_DEBUG_LEVEL and CMU_SMD_MONITORING_DEBUG_LEVEL to 9 in cmuserver.conf
Restart monitoring daemon, wait for few minutes. Then reboot one compute node, wait for few minutes and check if monitoring starts.
Please check these logs for errors:
On compute where secondary monitoring daemon is running:
On problematic node: