Symptom
We are observing following type of faults related to Atomic Counters:
moquery -c faultInfo -f 'fault.Inst.code*"F15[4|5]"' | grep code | sort | uniq
code : F1545
code : F1546
code : F1547
code : F1548
code : F1549
code : F1551
code : F1552
There's no impact on the network, but we need a fix.
Conditions
In 95% of all cases the alarms raised are false due to the nature of TEP-to-TEP, vPC-VIP-to-TEP, vPC-VIP-to-vPC-VIP forwarding behavior in the fabric.
But we need to have valid explanation of the root cause and find a fix.
Workaround
To clear these faults / not generate them:
Squelch faults. The steps to change the initial severity of the atomic counter faults to "squelched" are below:
* From the APIC GUI:
* Navigate to Fabric > Fabric Policies > Monitoring Policies > default > Fault severity assignment policies
* Click the pencil next to monitoring object
* Add a check mark by "Atomic Counter for a Trail"
* Add a check mark by "Atomic Counter for a Path from a VPC pair to a VPC pair"
* Submit
* Click the dropdown to the left of the pencil and select "Atomic Counter for a T
* Change the initial severity to squelched
* Click the dropdown to the left of the pencil and select "Atomic Counter for a Path from a VPC pair to a VPC pair"
* Change the initial severity to squelched
Further Problem Description
We need to have a better understanding of the fault and fix the issue
