The AMP CLI has some aliases that assist with SNMP walks.
For SNMP v2c:
# s2w (IP address of controller) (community string) (OID string) (additional flags)
Example:
# s2w 10.10.10.10 public nUserName
Example with flags:
# s2w 10.10.10.10 public nUserName -OX -Cc
-OX = translates hex values to mac if available
-Cc = makes sure to walk OIDs in increasing order (to escape loops)
For SNMP v3:
# s3w (AMP database ID for controller) (OID) (additional flags)
AMP database ID can be gotten from the URL of the controllers monitor or manage page. It will be the ending value of the URL. For instance, if my URL was: my.amp.name/ap_monitoring?id=15813
Then my ID is 15813.
This will merely confirm what data the controller passed to the AMP via SNMP. Pay attention to response time, a slow SNMP response from a walk is going to be the same when AMP makes the request. Also, if the table is large, you may want to time the amount of time it takes to get the complete response (In the case of AMP's request, the actual request is for the entire table, not just the single OID, so walking 'wlsxUserTable' would be closer to the actual. I had provided the individual OIDs as they get the specific data to look at out of the table).
If the SNMP walk yields the correct response, then the next step would be to open a support case. Provide the controller firmware version and the data collected via SNMP walk.