Why can the event EPS IN and EPS OUT statistics reported by Logger drop from a normal operating rate to goes to 0 randomly?
Logger 7.1.0 - Appliance (observed)
Can affect other versions.
The EPS IN and EPS OUT statistics of Logger can drop from their normal processing rate to 0 in a period of approximately 30 minutes.
This behaviour does not occur at a set (wall clock) time, but in general has been observed twice a day or so.
Full Storage Group(s).
For example, the issue has been observed when one storage group is almost at 100% for one storage group and the other at 85%:
Increase the size of storage group(s) or reduce retention policy.
As highlighted in the Logger Best Practices guide please note that to ensure better control of storage group retention and disk space utilization, do not allow storage group utilization to increase above 90%.
As storage groups near 99% utilization, Logger begins to run out of space to persist new events. The work (and time) required to free space to persist new events can cause them to wait (cache) periodically.