I am showing negative buffer misses on some of my devices. Anyone have any ideas?
Hi,
This information is pulled by SNMP requests.
Are these devices configured as SNMP Device Type in NPM ?
If you run a MIB walk on the OLD-CISCO-MEMORY-MIB on these devices do you get results?
http://www.oidview.com/mibs/9/OLD-CISCO-MEMORY-MIB.html
Check all the OIDs that end with *Miss from the above MIB.
Thanks,
Yann
Might not be the exact problem, but I've seen this happen to bandwidth reporting numbers when the interface is unabled to be fully polled for whatever reason. The DB inserts '-2' as the value. While the interface shows that it is operational, certain stats can't be "gotten" by Orion. In my case, it was due to polling values and deadlocks in the database, but has since been fixed by modifying the polling intervals.
Here is my MIB walk for one of the devices in question, looks fine to me. I also took a look at the buffers hourly and detail tables and there are no negative numbers.
Thanks for the details.
If you setup a wireshark trace, open the web console, go to Admin > Manage Nodes.
Select your device and click on More Actions > Poll Now.
Do you see it polling the *Miss OIDs ?
I did a quick capture, and it does hit the bufferElMiss OID, but not any of the others as far as I can tell.