11-24-2014 08:21 AM
I operate an VDX6720 VCS Cluster.
Since one week snmp access (from within the same network) is not longer possible.
Also if I try to remove and add snmp communities, the following error message appears:
iscsi-sw01_1(config-rbridge-id-1)# snmp-server community privsnmp
Warning: Cluster wide operation failed on Rbridge-id(s): 1. Succeeded on Rbridge-id(s): 2.
Rbridge-id(s): 1 Reason: %Error: Backend Get community failed
Network Operating System Version: 4.1.2a
I´m running the same config on another customer fabric without problems.
Do you have any ideas on this behaviour?
01-26-2015 10:46 AM - edited 01-26-2015 10:47 AM
There is a snmpd bug in all versions before 4.1.3 where the snmpd process is locked up by chance. You can see this with "show proc cpu" or login as root user to the switch and do a "ps ax".
As a root user you can kill the snmpd process with the kill command. This recovers the snmp process.
To fix this finally you need to upgrade to minimum 4.1.3 or later.