After trying to diagnose issues with support, buying all new servers with 15k RPM drive, loads of memory, CPU, etc... Still having the same problem. Chiefly, this:
Has anyone else experienced this problem? Usage graphs that just drop off for a period of time, then all of the sudden, come back.
NTA Storage is running on a separate DB than SQL, which is different than the Poller. All on 1Gb connections, with private connections between them (Poller takes in data on one NIC, sends to SQL and NTA on another).