[e-nsp] SNMP differences throwing me off

[Drew Weaver]

                Howdy.

I've been writing an application which will allow me to identify any device on my entire network no matter whether its connected to a Foundry, F10, Cisco, Watchguard, or Extreme device. The issue I'm running into is it seems that Extreme has some oddities when it comes to SNMP. (it could just be me, or it could be that the other vendors "went an extra mile to make things pleasant") I'm really not sure what the case is.

I've noticed that with other vendors' gear there are certain OIDs with which you can attach values to communities, like.. community at vlanid which will return specific info on that VLAN so that you don't have to snmpwalk an entire table and then parse it out (because its slow). I've also noticed that other OIDs (which again on other vendors gear) allow you to pass values to them for example snmpwalk -v1 -c community hostname atIfIndex.value (where value is a valid ID) should return a list, on the other vendors' boxes it does indeed, on my 6808 I find that I have to crawl the entire atifindex and then parse it...

I've also noticed that snmp itself is somewhat slower on these systems than on other vendors machines, has anyone found any ways to accelerate that?

Thanks for any advice.
-Drew
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://puck.nether.net/pipermail/extreme-nsp/attachments/20080201/8265ee69/attachment.html