We have a CA Netflow analysis system (NFA or RA) running version 9.3.3
We have recently reconfigured our routers, etc to use a different SNMP profile. In the Administration enable interfaces panel, the devices/interfaces sending netflow are showing up as using the previous profile. The only exception is when I go in and manually rediscover a device. Is there a way to manually start rediscovery on all devices?
We also use the Netvoyant application, and this must automatically do rediscoveries because all of the devices in there show they are using the new profile
I picked up your issue for this, but just saw this post out there so wanted to make sure the reply was here as well..
There no way through the GUI to force a rediscovery of all devices, however we can try to update mysql on each harvester to put the routers in an initialpoll state so that they start to snmp poll again.
To do this RDP to each Harvester and login to mysql with: mysql poller
Then run the follow two statements to update the database: update routers set state='InitialPoll'; update routers set updatedon=unix_timestamp();
Then recycle the poller service on each Harvester. This may force the devices to start to rediscover, it will take some time to process through all routers.
As Chris Walsh knows, I tried the above commands and it didn't work. Can't really remember if I got an error message or not. Chris gave me some other commands to run, but I never got a chance to schedule a change order to do the harvesters. So basically I ended up rediscovering devices manually.
These were the updated steps:
On the console run:
mysql reporter update routers set profileid=0;
This will set the SNMP profile to 0, telling NFA it needs to find a new profile ID.
Then on the Harvester server run:
mysql poller update routers set State='InitialPoll', stateretry=0, reboottime=0, updatedon=unix_timestamp(),discoverytime=0, profileid=0;
Then recycle the Netqos ReporterAnalyzer Manager service on the console server and the CA NFA Poller service on the Harvester server.