We recently upgraded our DCNM server to 6.3(1) and we've noticed that the server hosting the DCNM application is doing more IOPS than any server in our environment!

It seems to be doing 8,000 - 10,000 IOPS constantly.

In Windows you can see that postgres.exe process is doing an ton of reads from E:\Program Files\Cisco Systems\dcm\db\data\base.

Does anyone know why it's constantly doing so much IO?  Is this normal?

We simply use DCNM (aka Fabric Manager) to do some basic zoning on our MDS fabrics.

I've attached some screenshots to show what we are seeing.