low collector memory
I had a spike in my collector memory utilization - causing the collector to disable the instrumentation on 18 of our agents. This is the only spike I see of this magnitude. How can I find out what caused the spike? It occurred at 15:07 on 6-24-14
I would look at the bytes received measure of the collector to see if this was caused by agents sending more data than normal.
Usually agents send more data if there is spike in request count that results in more purepaths, also if there are errors then purepaths may be long and there can be many string data in exception messages, stack trace, logs etc.
Thanks for the suggestions!
There is no corresponding spike in bytes received, number of purepaths, purepaths analyzed or purepath length. Those measures all fall in line with our daily business cycle.
