automag928 ✭✭✭✭✭

My UCS is responding to hardware health, albeit only Temperature information. I set mine up per the instructions on the other thread, selecting the UCS Manager credentials under the additional monitoring options of the node in NPM.

Yep I thought of that, stat collection is 15, rediscovery is 30, poll interval is 120 in the volumes DB

Yes there sure is - you can mark an interface "unplugged" rather than down -

Take a look at this thread here and it should solve your problems - NPM Alert on Node UP (Ping) SNMP DOWN (unknown / unresponsive) Bascially you are collecting cpu / ram via snmp - and the query goes out to see if you've gotten data - and this will alert you if ICMP is still responding, but snmp data collection is failed.…

Based on my experience traps and polls are separate things. You can receive traps or poll devices, or do both and they are configured differently on the end devices I've got experience with, i.e. Cisco networking gear, Juniper stuff, Windows and linux servers etc.

Just curious and I know it doesnt help but what version OA administrator are you running? I've got anywhere from 3.50 to 4.01 and my CPUs are reporting normal.

Could you use the complex conditions to create two blocks for evaluation of trigger actions? Like your primary section would be panSysHAState = passive and then your second section would be interface not down? So the first condition will get evaluated, the HAState is passive, then move to the second section, and make that…

cvachovecj‌ - has anything come about from this?

Yikes - I'd be tempted to start it up clean install, fresh DB everything. Set it up with custom properties , groupings etc, re-learn all the correct interfaces. Thats a tough call.

I should also add, alerting is also affected - not just switch config visibility. The individual teams don't want to see other teams' alerts.

Are you monitoring all of those interfaces in NPM? I would think you could create a report from NPM with traffic stats on those ports, rather than UDT

Just out of curiosity - have you tried under the right hand top down (the three lines with the down arrow) , and hit show swql - then goto http://yoururlforyourswpage/Orion/Admin/swis.aspx and paste that swql query in there and see the results you. You can also work with the SWQL query and section out the different parts…

I put in a feature request to have this polling issue fixed -Fix UDT polling for Cisco Nexus FEX ports

adnan.ibric‌ - you should just be able to configure your 3rd party software to send snmp traps to the IP address of your solar winds snmp trap receiver. To display the trap, I'm assuming you would need that node that generated the trap to be in solar winds, and then solar winds could display it referencing that node.…

Originally I was messing with the reports but from the solarwinds report scheduler it looked like I was hamstringed into only once daily - hence why I went to alerts...but then the light bulb went off, and I remembered I could change it via windows scheduler. Thank you!!

We've got our NTA db a seperate vm. Its a db server, and depending on your netflow's config'd you may have a good amount of resource usage. Same with the web use...We've got an additional webserver but it doesnt take alot of traffic. If they are all VMs, whats the downside to assigning them their own machines?

christopherbell‌ - I've spent a considerable amount of time trying to get UCS chassis monitoring. Its is multiple level setup. You add the UCS Manager GUI IP and monitor it via ICMP and SNMP, as well as checking the box for inputting the UCS Credentials, where you fill in the port # and login information for the UCS…

Do you have CIMC IPs assigned to the blades on the UCS?

I created a custom property called "owner" and while this is more group specific in my case, you could apply it to individual users as well. With the custom property defined on all of the devices, you would then modify the view of the dashboard elements to Filter Nodes (SQL) and Owner = "xyz user" Depending on how many…

Go into license management and deactivate the SW licenses, then uninstall SW or reformat or whatever.

Just curious if anything came about from this...I'm in the same boat - looking to monitor ssl certs on the ACE. The command run to show them now is "show crypto cert all", but some certs dont have a website tied to them so the SSL cert monitor for sam wont work.

Was there any fix to this? I'm running NCM 7.1.1 and still cannot clear transfer statues that are created by other users...

I'm pretty sure this is not available, at least in my 10.7 version of NPM. Not sure if it was updated in 11, but I'm guessing no.

For kicks , I deleted one of the two dl380s that was no longer polling hw health, and re-added it - still no hw health polling option in list resources.

Oh, and the hw details section on those two nodes shows this: Hardware Status Undefined Message Hardware polling failed: Object reference not set to an instance of an object. Manufacturer HP

pres1dentnix0n‌ - to do some troubleshooting, since I don't know the alcatel's that well, can you edit that specific poller, then on the first tab of "define properties" select the test node of one of the alcatels you would expect to get results from then hit next. On the next page you should see the defined pollers and…

I group by several different custom properties, such as location, ownership, device type (ie, server, firewall, switch, load balancer, etc)

Thats interesting. I typically don't do scheduled auto-discoveries, typically I create them and then run them immediately. So as a test, I went and looked at one of my previous manual ones - added in a scheduled discovery for 3 weeks from now, then went back and hit run manual and got the same error. So going back to the…

Did you ever figure this out?

I've had a case open with Solarwinds and Cisco development for our issue. UDT wasn't populating the mac/IPs for ALL ports, but only one or two vlans. After quite a bit of troubleshooting and research, we noticed this was specific to FEX ports off of our nexus. We were seeing only the ports that had vlans on the line cards…