I'm chasing down some storage latency issues reported by our ESXi hosts. All of my hosts (~60 across 5 clusters) report the same storage latency warnings at the same time each day: 03:00. The warnings last for 2-3 minutes, then performance returns to normal.
I logged into our HDvM to see if similar performance problems were observed on that end, and found that the same latency spikes can be seen from the VSP. I also noticed that there is a scheduled "Health Check" tasks that runs at 03:00 each week, and lasts 2-3 minutes.
Here's the question: can someone explain (or point me to some documentation for) the Health Check task? And has anyone else observed similar behavior on hosts during the Health Check? The timing is suspicious to me.