How to change Rediscovery Interval per node/interface?

[AndyCoates]

Hi,

How can we change the rediscovery interval per node/interface?

Each node and interface object has a default value set, but it doesn't seem to be editable anywhere.  We could probably change it via the database, but that would probably mean a NPM restart which we'd have to schedule and can be awkward at times. 

Is there any way of changing it on-the-fly?

Thanks,

Andy.

[pacetti]

AndyCoates,

You can edit the Default Rediscovery Interval and all other polling intervals on the Orion Polling Settings view, as follows:

1. Log in to the web console using an account with administrative privileges.
2. Click Settings in the top right corner.
3. Click Polling Settings in the Settings group.
4. Edit the Default Rediscovery Interval and any other polling intervals, as desired.

HTH,

[AndyCoates]

As the topic states, I need to change the interval per node/interface/volume etc, not the default interval.

The only way this is possible appears to be editing the field directly via the database, which unfortunately will mean a NPM restart.  For those curious, each node/interface/volume table has a RediscoveryInterval column that should have the default interval already entered.  To change it, you'd need to stop Orion services and change that field, then restart the services.

Not an elegant method, but hopefully in future releases Orion will have this option somewhere on the web GUI (similar to how Polling Interval is editable).

Andy.

[ChrystalT]

If you are wanting to do a lot of them for the same interval, why not use the network sonar discovery?  You set up how often it runs a new discovery.

[AndyCoates]

Unless I'm confusing what you're saying, that is for discovering new nodes?

Just to clarify, we have the nodes in our database, we just want to change the 'rediscovery' (when it goes off and scans for interface re-indexing and such) to a different value than the default for specific nodes/interfaces.

Andy.

[ChrystalT]

It will discover any new interfaces for you if the nodes are already in NPM.  Then you can choose which ones to import.  You can scan by specific IPs, subnet or supernet, etc.  It scans the same one every so often (whatever you set it for) and will let you know when there are new interfaces/volumes.

[AndyCoates]

Ah I get where you're going with this - I hadn't thought of using it in that sense.

Unfortunately it's the opposite of what we're trying to achieve.  We *don't* want to rediscover anything, or at least set the automatic rediscovery NPM does to a very high value for certain nodes/interfaces.  There appears to be a bug with some old Cisco Catalyst switches since we started polling them (SNMP related), so we want to reduce SNMP traffic as much as possible to these hosts and taking off or increasing rediscovery interval would help.

Thanks,

Andy.

[ChrystalT]

Oh I understand now.  I can't think of a way to do this off the top of my head.  Sorry.