Help with performance policy on Exchange 2016

[Sup3rFly]

I have a Nimble CS300.  I recently built a new Exchange 2016 VM running Server 2016. With only a few mailboxes on the server, I'm getting disk latency errors from Veeam ONE Monitor on my logs volume.  I rarely see this email alerts and never on my old Exchange 2013 VM or SQL VMs.   Veeam ONE is showing latency between 40-70 Milliseconds, about 20 alerts a day.

I am running ESXi 6.5.  Exchange is on VMFS6 datastores.  I used the VMWare ESX 5 performance policy for OS, Database and Logs on my 2016 server.  Should  I use the out of box Exchange 2010 Database and Exchange-Log policies?  

Thanx!

[Sup3rFly · Accepted solution]

Changed Logs to Exchange-Log policy with no caching and so far no errors while I'm migrating a bunch of mailboxes.  Hopefully that was it!

[Nick_Dyer]

This shouldn't make a lot of difference.... Do you have the VMware Cross-Stack Analytics configured? If so, you should use that to troubleshoot and cross-reference latency figures from Veeam One.

A recent example is that a customer was seeing high latency in Veeam One and was blaming Nimble (even though it was an AFA), however the cross-stack analytics pinpointed easily that the latency was due to high CPU Ready on an ESX server, due to a noisey VM that had been migrated a day or two previously. The array was not the culprit.