2014-08-28 09:47 PM
I have purchased a FAS8020 2 node CDOT cluster. Only using it for SMB3 for HyperV.
I have created 2 aggregates, one on each shelf.
Do I need 2 SVMs with their volumes in each separate aggregate to spread the cpu load or will one SVM still spread the load across both nodes?
Solved! SEE THE SOLUTION
2014-08-28 09:54 PM
SVM does not perform any “load spreading” automatically. You can have volumes from multiple aggregates in the same SVM; but assigning your workload to these volumes is up to you as administrator.
2014-08-28 09:58 PM
Sorry I should have explained more.
I create a single SVM, with 2 volumes in separate aggregates assigned to different nodes. When a CPU intensive task like dedupe runs, does it use the cpu of the node hosting the aggregate?