08-17-2015 01:03 PM
We have an OC setup running version 6.9.4. We notice a lot of time we are missing end host IP information in End Host Report (SPM). At this point, we know for sure the firewalls which holds all the ARP information is in our NetMRI system. Yet, we can't find some Arp entries in some cases (using FindIT), but sometimes we can for other Arps (Even thougth they are from the same firewall).
4 things I can think in our envrionment setup that I think can affect this:
1. We don't have "Refresh device caches before collecting switch port data" and "Probe for open ports" in our group definition.
2. Our ARP Collection Priority is selected as SNMP.
3. Also, we don't add discovery ranges, instead we add all devices statically. Does this also affect our chance finding end host IPs?
4. Our SPM polling cycle is 24 hours.
Or this is just NetMRI's limitation overall.
Solved! Go to Solution.
08-21-2015 11:27 AM
I would suggest opening a case so they can deep dive to see what the issue is.
09-08-2015 09:33 PM
This might be worth testing on some devices (small IP range) to see if it helps... It has been awhile, but the last time I saw this issue, it was resolved by adding IP address ranges of the end hosts into NetMRI, but adding them as "exclude from management". This allowed NetMRI to identify and create the end hosts as objects, but witholds NetMRI from trying to "manage" those IP addresses with the sort of interaction netMRI conducts with routers, switches, etc.
This is something I often recommend to customers to do in any case to help ensure NetMRI is efficient in the network.
If you're willing to share, may I ask why you enter the managed device IP's statically rather than entering some seeds and allowing NetMRI do discover?
09-28-2015 05:45 PM
Yes, that does the trick. By making network range as "exclude from management", we are seeing more end hosts showing up in the end host reports.
I have a follow up question: for end hosts fall in the "exclude from management" range, is there absolute 0 network interacation between NetMRI and end hosts? Or there is some minimum interaction (like ping)?
It is just a question we have.