Analysis Services low thread count and CPU usage on one server, high on another
We have an Analysis Services instance that runs fast on one server (environment A) and slow on another (environment B). B is more than twice as slow as A: at least 7 hours slower.
Performance monitor shows that environment A is using 95%+ CPU and creates 60+ Processing Pool threads. Environment B has 2 threads for the Processing pool and sits at 20% CPU.
The msmdsrv.ini files are identical, except for the disk locations. MaxThreads is set to 0 (automatic).
On identical hardware with identical msmdsrv config files, is there anything else that changes how many threads SSAS processing will use?
For additional context:
Both have identical CPU hardware: 2 sockets, 8 vCPUs. The databases they read from are identical. We’ve connected cubes from one server to the warehouses on the other, and processing is only slow on environment B regardless of where the SQL database is.
RAM allocation is the same, there’s plenty of space left.
Disk usage is normal, the drives are spread out over different drives in the same configuration.
We have an Analysis Services instance that runs fast on one server (environment A) and slow on another (environment B). B is more than twice as slow as A: at least 7 hours slower. Performance monitor shows that environment A is using 95%+ CPU and creates 60+ Processing Pool threads. Environment B has 2 threads for the Processing pool and sits at 20% CPU. The msmdsrv.ini files are identical, except for the disk locations. MaxThreads is set to 0 (automatic). On identical hardware with identical msmdsrv config files, is there anything else that changes how many threads SSAS processing will use? For additional context:Both have identical CPU hardware: 2 sockets, 8 vCPUs. The databases they read from are identical. We’ve connected cubes from one server to the warehouses on the other, and processing is only slow on environment B regardless of where the SQL database is.RAM allocation is the same, there’s plenty of space left.Disk usage is normal, the drives are spread out over different drives in the same configuration. Read More
