I don't have experience with SMVI 2.0 yet (soon), but on SMVI 1.0 we have 15 (very large 24CPU, 256GB RAM) hosts with 7 NFS datastore and about 425 VMs. I did notice that if I create a single schedule to perform the snapshots, I would get timeouts on disk on VMs and a lot of alerts about high latency. I created separate schedules for each datastore and staggered them by 10 minutes--all the latency issues and timeouts disappeared. This surprised me since I have filer based daily snapshots which all occur at midnight as a backup (in case SMVI is not functioning I still want snapshots). Those never caused timeouts, but I suspect there must be logic to handle that workload serially.
