I am rolling out NPM. What are the best out of box alerts for Network Monitoring? I have some ideas but wanted to see what others have done in their rollouts? Thanks
Most of ours are custom. Out of the box are basic and work great. However, it is only might to give a real high level view. I'll give you an example, Reboot alerts, If you received an alert that just says "Server has rebooted" how helpful would that be to you? It has no substance. But now if you had an alert that give you everything, lets say its a windows device so it even gave you event ID's from the event viewer now this would give you a lead on what the problem could be. It makes it more efficient. The good part about alerts is that they can be twists modeled and shaped to work just the way you want them too. But I would say if you want anything more than just high level view alerts than I would suggest customizing them to alert based on your needs. I hope this helps! If not, well I tried... haha! Good luck.
Thank you, I am tracking on custom alerts for System Nodes. However I am looking for the High level alerts for the Network Switches, Routers, Vlan, VFrs, Interfaces and F5s?
In that case I don't see why the out of box wouldn't work in your case. Maybe you'd might want to duplicate and edit them just so you can make minor modifications to them so that they can alert properly. By default I believe that the out of the box alerts should be able to work on the remaining items you've listed. Have you tested them by chance? Are they working? Any specific information they aren't providing? Just a few things to look at. It would help me help you a little bit more.
Absolutely, yes, I am meeting with the Network guys next week to get an idea what they want to monitor. I let you know as I defined our alerts. Thanks
You can't modify the 'out-of-the-box' alerts. So you'll have to copy them.
But they are good to give you an idea on what alerts you can configure but a simple one like node down, if you use the out-of-the-box alert, you'll get an email with little info. I would custom write your email alerts.
Depending on how big your company is
My methodology has always been make alerts actionable. Upper management wants "proactive" or transnational monitoring to see if a system is slow. You have to walk before you can run and you have to crawl before you can walk. So start out with
1) Node Up/Down. If proven they can handle this do
2) Server disk space full
3) CPU/Memory usage alerts
4) Application Stack
If people aren't responding to alerts its just noise. There has to be accountability for alerts for people to troubleshoot it.
