rharland2012

Not positive on that - but I saw on your port config that the max datagram size was the same as your max header, and that sounded off. If you're getting good flows, I'd watch your traffic to verify sample accuracy and change it only if need be. sFlow, in a perfect world, is a low-touch protocol.

Okay. Are you trying to create the recording on your server, your PC, or some special-purpose box? Where are the other successful recordings sourced from?

Don't be afraid to investigate the building blocks of the application monitoring templates, either. There are many monitors I use today that are just smaller, more digestible subsets of the full-blown templates with only the pieces our users want/need to see in the finished product, which is usually a copy of the vanilla…

If you're running the appropriate version of PS on your polling box and verify credentials to access your O365 environment, then yes.

Even more elegant!

That's an unusual issue! I've had problems before, but it always boiled down to permissions of some kind. With any luck, some of our fellow forum members might be able to contribute. I assume this is a domain machine - or at least identical from a topology standpoint to other servers you're successfully monitoring.

Is the SNMP service running on the vCenter box? Are you able to successfully authenticate via SNMP from some other utility to test that it actually answers?

Don't leave me hanging, dude - what was the answer?

Absolutely it can. What I do is create a copy of the alert 'Alert me when a component goes down' and put conditional statements in it to point at the node/component in question. Works well.

This almost sounds like a user definition issue. I remember this was missing for me as well. In Admin, manage your user and go into Server and Application Monitor settings. Make sure the RTPE is allowed under your profile.

That's a strange one. Do you have existing nodes in your DB that populated normally, or is this a new installation? Either way, I would suggest opening a case with support. The timeout piece is weird - and if you're on the same segment (or just a hop or two away), then timeouts shouldn't be occurring, I don't think. As a…

Strange. I do seem to remember some old threads discussing something similar to this. I would suggest opening a case with SW support - and in the meantime, perhaps you'll get some more feedback here. I think the limitation may be that since the thin APs are not managed nodes per se it may be difficult to customize the…

Alex, We lit up DPA about a year ago, and here are my thoughts on your questions. 1. It does officially have its own DB (the repository). I placed this DB on the same SQL instance as our Orion DB with no ill effects, but we're a sub-10000-element single-poller setup for now, so if you're scaling up your mileage may vary.…

Corey, did you have success making this change? We see some of this with our Riverbed boxes as well.

What OS are the problem servers? Are they the same as successfully polled agent-based nodes? Are there any roles/features unique to these two problem servers?

Even better! balav​, you should definitely have what you need now.

Of course....duh...thanks!

nei4352‌ - I would never do such a thing! Well, hardly ever.

I wanted to see if anyone would come up with something good!

I'm working on it....

No, it doesn't have multiple IP addresses, and the address it does have is not equal to any router address on our network. It seems to be an NPM- specific issue, as next hops report normally from the router itself.

Is there anything in the Fortinet MIB that would reveal this information accurately? Seems like a UnDP might be a good option here.

Weird...I haven't seen that issue. Sorry to hear you're having problems - kind of a drag.

No problem at all - if you're still having issues polling for hardware health/etc., just make sure the interface you're talking to via SNMP has SNMP access enabled in the GUI. You find this in the Network>Interfaces menu via GUI.

You just have to log into thwack to be able to reply, etc. That's a great question - I'm sorry I glossed over what you were truly after! I haven't yet attempted to change node status with a UDP alert. People *have* asked for this feature in the past, I believe. This would be a great feature request! I'll dig a little bit…

Okay, great. Do ALL nodes show down, or do some show as up?

Have you attempted to search on all 255s, instead of the subnet's logical broadcast? You may be surprised at what you see.

None of the tutorials or other supporting documentation in the Library & Support section for NCM helped you?

I'm not aware of granular perms settings for UDT only. I would suggest either opening a case with support to get validation from SW or perhaps waiting a bit to see if richer minds than mine on thwack can help you accomplish the goal. There may well be some method to accomplish this from the SQL side.

Ah, I see. Thank you.