Hello,
We keep getting an alarm about High Operational DB Usage, The alarm appears about once an hour and is basically the following.
"ISE Monitor Node(s) is about to exceed the maximum amount of allocated RADIUS disk space.".
Looking at the detail of the alarm, I see that the Tacacs Data threshold_space = 95 GB, used_space = 0 GB, which I would expect since we are not using TACACS. the other description that is causing the alarm is Radius Data threshold_space = 143 GB, used_space = 119 GB. we currently have dropped the number of days that we keep operational data from 90 to 60 days, but when we did that, almost an hour later we started getting the alarm again. I see in the log where it appears to be purging the RADIUS data older than xx Month year (04-Feb-17) is occurring and succeeds, that is the 60 day limit that we set.
What data is kept in the RADIUS DB, we keep getting an alarm about mis configured network devices, 12929 NAS sends RADIUS accounting update messages too frequently, and that is a very high number, 50K attempts at times, but I have been told that this is a COSMETIC error, nothing really wrong with that message by TAC. we have also tried to limit the number of accounting updates with no real change.
My concern is that if I keep dropping the data retention period say from 60 to 30 days, I am still going to fill up the RADIUS DB. How can I tell what type of data is really overloading the RADIUS DB so that we do not get these cosmetic alarms as it was put to us, and address the real issue of what is filling up our RADIUS DB. We have what I would consider a medium size deployment with what ISE See's is about 14450 total endpoints.
Thanks,
Sam
