Hi @JSmith ,
Assuming there's nothing corrupt in the system, resources are the main reason for slowness.
You state the scan time is about an hour forty five behind but is that how long it is actually taking to scan or is that the last timestamp. If it's just the timestamp and if the polls are indeed 15 minutes between each other, perhaps it's how fast the system is able to write to disk.
Since you state it's a powerhouse, it must be a VM as DCE hardware is somewhat generic. If the VMWare host's disks are remote...the VM host uses a NAS for data storage, it could simply be the IOPS / writes are too slow.
You've also got a lot of devices. 2000 or so devices using SNMP version 1 is a lot different than 2000 devices using SNMP version 3. I've been asking for better numbers for SNMP version 3 as although we recommend specific resources based on number of devices, I believe we need to make different recommendations based on protocol. There's a ton more traffic and processing needed for SNMP V3 than there is for V1.
You may want to have a look at the following performance troubleshooting guide in any case:
https://dcimsupport.ecostruxureit.com/hc/en-us/articles/360039287893-DCE-performance-troubleshooting...
And here's what we recommend for sizing. There was an older one that showed lower resources and again keep in mind, these numbers are based on SNMP V1:
https://dcimsupport.ecostruxureit.com/hc/en-us/articles/360039287913-DCE-sizing-guide
Thanks,
Steve