I'm seeing a few devices stop responding to polls and then immediately (within the same second) the alarm is cleared. I see no activity where the device was put into maintenance or hibernation, nor do I see discovery information regarding those devices during those times. I've contacted the server owners and they report no log entries around those times, and the network group says they have no log entries during those times. As far as we can tell these are false alarms. I'm unsure why we are getting these alarms and how to prevent them from happening. Any suggestions?
What is ping latency from the spectrum server to the end device ?Try to increase the DCM timeout and check it once ..
Thank you for your reply. From the Spectroserver to one of the devices a tracert shows 6ms. The DCM Timeout is set to 3000ms.
Are these modeled with v3? If so, do they share the same engineid as another device? If so, Spectrum currently does not support devices with duplicate engine ids. If not, I would encourage you to run a sniffer trace and create a case for us to investigate.
These are SNMPv1. I think I'll open a case with CA. Thank you for your reply.