I fighting still weeks with the nvidia support about issues with the nvidia license server.
I use the latest version from november 2020.
I have a primary node and also a failover/backup node. Both installed on windows server 2016.
I can see the replication works between primary and backup license sever.
But something is wrong here…
The database files grows on the secondary site each day. The database file on the primary is not larger than 10MB. On the secondary site it is already 120MB after one week. I reinstalled the license server on the secondary site multiple times.
I can also see in the log files of the secondary following errors:
17:34:41,213 INFO Scheduled-pool-9-thread-7 [sync-session=6YDPLL5LK8KM] Return for feature GRID-Virtual-PC 2.0, count=-1 failed.
17:34:41,213 INFO Scheduled-pool-9-thread-7 [sync-session=6YDPLL5LK8KM] Return for feature GRID-Virtual-PC 2.0, count=-1 failed.
Can anyone of you confirm that they have the same issue?
I’ve not experienced this before, however all my License Servers are run on Server 2019.
A few options to investigate …
Try a clean install of Server 2019 (fully patched)
Try a different version of Java / OpenJDK
Try using Linux for the License Server
Are you running anything else on these License Servers apart from the NVIDIA Licensing? Any other Server Roles etc?
Just looking for anything at this stage that would give an indication of what’s causing the issue, as NVIDIA Support are pretty good and the License Server components are fairly basic.
Thanks for checking.
It looks like you don‘t use the lastet release of the license server from november 2020?
You also use oracle java version? We switched to openjdk because of oracle license policy.