rharland2012

I'll defer to other thwacksters or SW personnel to give the official answer, but my feeling is the same as yours - that would be a no.

Were you able to glean anything useful out of the cap? I'm out of ideas at this point.

Let me know if you didn't get that pcap. This is a real puzzler! I can't see a thing wrong with the config on the Fortigate - I've opened a ticket with them, investigated the same stuff, and we found the same thing - we see the packets exiting the Fortigate, and we see them hitting the SW box. Strange.

Yes, it is. Are you running VDOMs?

I no longer get the squeezed top-left thing, but I abandoned the use of the rotating maps since about half the time our nodes don't populate correctly and I'm left with a blank global map. It might be Chrome-related or something else, but it's just not dependable and there's precious little info out there on a permanent…

rob.hock, what's the official word on this issue? Thanks for any info.

Thank you, Rob. Let me know if there's anything I can do (within non-invasive reason) to help work toward resolution. I'm really looking forward to showing off these NOC views!

Check this out: NetFlow setup for 3750x with C3KX-SM-10G | Advanced Data

Understood. Color me corrected.

scottyplo​ - DM me your mailing address, t-shirt choice and size and I'll buy you one if you want.

Not a Pure shop here, but I'm almost positive their boxes have some SNMP usability. Are you encountering an error when trying to add via SNMP?

When editing the Email/page action for the alert, expand the 'Message' tab and scrub out the text you don't wish to retain. Choose 'Insert Variable', select 'Node' under the 'Show Variables For' dropdown, and you should be presented with choices there for populating the message with the node name. Then, craft the message…

Okay. I don't know how to accomplish that either! With any luck, you'll get your method from one of the many skilled forum members around. Following this so I can find the method as well.

What I did was spin up a separate VM to act solely as the FSDB. This box is separate from my NPM/NTA front-end box and separate from my SQL box. It's very, very simple, low-touch, and has worked flawlessly for me since day one. I understand that being able to spin up a box just for this purpose is a luxury, but it has…

Here's where I would start. First, define whether or not any firewall policies are in play between the poller and the computer you're trying to add. If there are, you'll need to allow the appropriate communication between the two. Based on your original message, it sounds as if ICMP is allowed - but that doesn't mean that…

Okay, so you already set up a group with 'thousands of interfaces'. That was also a lot of work, sounds like. Are there any commonalities to the nodes to which these interfaces belong? Do the nodes contain custom properties or other properties are the same? If so, you may be able to build an accurate alert based on high…

Well, it depends. If you have NTA to leverage and are collecting info about your traffic, then you can filter out IP ranges/etc. via the left column's Flow Navigator and would then save that view to the menubar so I could view it whenever I wanted. If I wanted to do it in NPM, I would copy the view where the TopXX chart…

Yes, I'm sorry - should have clarified that. Your firewall egress to your providers. Since this is a polling situation, your poller will initiate the connection - and with most stateful firewalls, you won't need to massage any incoming policies. If you're doing traps from your providers, then you'll have to add incoming…

Yeah, sorry about that....i'm a bonehead. Anyway....open the page with cache flushed, etc....demouser goes quickly to 'guest29' and spins for posterity. No joy. IE10 win764 a:a.

Can you ping the IP of eth1 from your comp?

Not sure about the hosts false alert, but the VM alert I assume is looking at some uptime counter on the host/cluster. Are you using VMAN to monitor clusters or separate hosts?

I was not aware of that - strange. I can't find anything in the docs on such a limitation after a quick readthrough. chrispaap​, can you confirm export limits as jwester​ mentions above?

Customers love it when the Head Geek refers to their hardware as a POS.

It wasn't my intention to take anything off-track - just correcting your assertion regarding the possibility of getting traps from a provider/carrier's device to NPM. Best of luck to you.

Can you provide a screencap of the error you're seeing on an example problem node?

Okay, good. Are you using net-snmp to poll these processes?

Hm, that's really weird. I don't see it on my ICMP node details view or via node management. I even double-checked the column selections as well - just shows up as blank if I enable the 'Overall Hardware Status (Node) column. What version of NPM are you on currently?

nProbe won't solve this problem for you, since you still need a device that will create and export the flows themselves. In other words, if your switch was netflow/sflow-capable, for example, you would use that switch to export the flow to your Solarwinds collector.

The VM is monitored via VMware polling of our vCenter and via WMI. Thanks for your suggestion.

Thank you - and thanks for the nudge to do it.