01-10-2019 12:49 AM
I have a situation at one of my customers where periodically (a few times a week) the nbstsserv process crashes.
They have a clustered master server (with VCS) and this then causes the nbu_group resource to fail its monitoring process, which in turn brings the entire thing down, and the up again on another node.
I have found an eTrack (3941164) which is fixed in 8.1.2, but only refers to 'nbstlutil cancel" commands causing the problem.
We get this even when NO cancels are being issued.
The customer does do a lot of SLP reconfig's (due to sp1ace constraints on their multipule target devices).
01-10-2019 12:50 AM
They are using NBU 8.1.1 on RedHat Linux (6.9) with InfoScale 7.3.1
01-10-2019 12:49 PM - edited 01-10-2019 12:50 PM
We are on solaris not linux, but we had our SLP duplications "hang" recently, I had to kill and restart the nbstserv to get them going. I make changes to my SLP config all the time, but I do not cancel anything, I just let them take effect the next time it processes. This is the first issue I have had in several years, with this daemon.
You may want to stop using the cancel option, I was told it can cause issues. Looks like you confirmed it!
If you find a solution, please post it!