Tag not monitored by Microsoft.
Microsoft HPC Pack 2019 - Invalid node groups issue
Intermittently, while submitting jobs, a pop-up comes which says "Invalid Node group <Name.>"
After that, HPC doesn't accept any job, until we restart HPC Management & HPC Scheduler services.
After service restart, everything becomes normal.
Below logs seems to be suspicious to me from HPC Management trace logs:
[SchedulerNodeService] Exception:.Microsoft.SystemDefinitionModel.InstanceCacheLoadException: Failed to load instance id 00000000-0000-0000-0000-000000000000 in change 00000000-0000-0000-0000-000000000000, revision 0 from the store. .. at Microsoft.SystemDefinitionModel.InstanceSpace.IdentifiableInstanceCache.GetInstance(InstanceIdentifier instanceId, ModelQuery query).. at Microsoft.SystemDefinitionModel.ModelQuery.GetInstance(Guid instanceId).. at Microsoft.SystemDefinitionModel.ModelQuery.GetRootInstance(Boolean createIfMissing).. at Microsoft.SystemDefinitionModel.ModelQuery.FindInstance(String xpath).. at Microsoft.ComputeCluster.Management.SchedulerNodeService.GetCluster()