This post was originally published on this site
Recently I was in a situation where a customer needed to capture esxtop data during a performance impacting event, however the cause of the event was unknown. What this resulted in was hours and hours of testing (e.g. throwing large amounts of traffic & compute load against some VMs), hoping that the event would trigger.
There were a few downsides to this approach, most notably:
Significant investment in time, most of which was spent watching nothing happen; Not knowing what datapoints we needed to capture to isolate the problem (e.g. host, storage, network, application metrics, etc.)l The risk of missing something during an event (which may last 30 seconds … there is a lot of data to look at in that time, especially if you don’t know what you are looking for).
Just like how PING is one of the most useful tools when troubleshooting network issues, esxtop is up there with displaying all sorts of data about your ESXi host, including VM performance and networks stats.
CAPTURING THE DATA
We all agreed the esxtop data would be ideal, but screen captures wouldn’t cut it. That’s where esxtop batch mode comes in. If you aren’t familiar with it, batch mode
Want to learn more? Check out the source post!