Hello all,
I'm running NSX 6.4.1 build 8599035 and vcenter 6.0.0. 5112527, Yesterday I re-deployed Trend Micro Deep Security SVM (DSVA) to 31 hosts, except for one host, the SVM (DSVA) wont re-deploy. I've tried rebooting the host (twice), I also tried putting the host in & out of maintenance mode --it powers off the guest introspection GI and powers it back on, but won't re-deploy the Trend SVM. I also tried clicking on the resolve button but nothing happens, I thought EAM would redeploy it.
Any ideas, which logs should I check?
Thanks,
-n00b-
Hi n00ban00ba,
Is there still a Trend Micro SVM on the problematic host? If so, the resolve button for the service deployment won't attempt a re-deploy if the appliance still exists and looks functional. You can try to delete it from disk an retry the resolve operation.
If that doesn't work or the SVM doesn't exist, you can try a full 'unprep' of the ESXi host in question. To do this, the following must be done:
1. Put the host in maintenance mode.
2. Remove the host from the cluster. This will trigger the removal of all NSX VIBs, including the endpoint MUX VIBs, GI VMs etc. A reboot shouldn't be required as you are running 6.4.x.
3. Check in NSX to ensure the host is no longer a member of the cluster (i.e. not showing up as 'Not Ready' or in error state).
4. If any SVMs still exist, delete them manually at this point.
5. Add the host back to the cluster. This should trigger MUX re-install and re-deploy both the GI VM and Trend SVM.
Regards,
Mike
Hello, the flowchart on how to troubleshoot service deployment i.e. Trend Micro in your case, is described in this link: Service Deployment Fails with OVF/VIB Not Accessible Error
In terms of logs, you can start from eam.log, NSX Manager log, DNS/FQDN/NTP of NSX/vCenter/ESXi
You can also check the vCenter events, make sure there are enough resources (CPU/Memory/Storage) in the target ESXi host
Hi n00ban00ba,
Is there still a Trend Micro SVM on the problematic host? If so, the resolve button for the service deployment won't attempt a re-deploy if the appliance still exists and looks functional. You can try to delete it from disk an retry the resolve operation.
If that doesn't work or the SVM doesn't exist, you can try a full 'unprep' of the ESXi host in question. To do this, the following must be done:
1. Put the host in maintenance mode.
2. Remove the host from the cluster. This will trigger the removal of all NSX VIBs, including the endpoint MUX VIBs, GI VMs etc. A reboot shouldn't be required as you are running 6.4.x.
3. Check in NSX to ensure the host is no longer a member of the cluster (i.e. not showing up as 'Not Ready' or in error state).
4. If any SVMs still exist, delete them manually at this point.
5. Add the host back to the cluster. This should trigger MUX re-install and re-deploy both the GI VM and Trend SVM.
Regards,
Mike
Woohoo this worked and I didn't have to remove the host from the distributed swtiches.
Thanks Mike!!!
Glad to hear that did the trick!