mcbridea

Unmanage the VLAN IF and ignore the warnings in the NTA events resource.

Hi freemen, Yes - what you described is normal behavior. I would recommend monitoring only the ports/apps you are concerned with. Monitoring all does make a very large number of ports and can cause you to have a very large NetFlow db. The default list is one we have instituted as commonly applied ports. These can be…

The Orion NPM Administrators Guide calls out the requirement in the System Requirements section. http://www.solarwinds.com/support/orion/orionDoc.aspx There are a number of references out on the web regarding the benefits, but they all are written in DBA talk. For Orion, CLR allows for improved performance. Andy

Hi sunday, Alerts work by querying the database for a condition to be true. A trigger condition of Node Status is Equal to Down will trigger if all nodes are marked in the database as down when he query is executed. 100 nodes down will = 100 alerts. The database is updated for node status when the status polling is…

Hmmmm. maybe change the property name to IBM and make it a yes/no type.

I'll have a developer comment.

ip route cache flow is only valid on the main interface and was superseded by the ip flow ingress|egress commands to allow v5 directional exports and sub interface specific exports. v9 marks the export PDUs with a direction indicator, v5 does not but it still exports ingress and egress.

Check the bandwidth statement in the device config. Is NPM showing the correct bandwidth for that IF type?

You can set it up as you would on a physical server.

Probably not.

The "Ten Minutes" in the title is not a good choice. There is a lot of content and good answers. I don't know why the author felt it necessary to go the "5 minute abs" path. The book is much deeper than the title suggests.

You can remove AS zero in NetFlow Settings > Manage Autonomous Systems.

I checked with one of our application engineers. He does not recommend this feature as the most common problem with Orion performance is the performance of the SQL server. Also be aware that we have not QA tested our products with SQL 2008 data compression yet.

If you want I can look over the pcap. Sent to andy.mcbride@solarwinds.com

http://www.solarwinds.com/support/Orion/docs/OrionNPMAdministratorGuide.pdf

Look at the MAC addresses in Switch Port Mapper. The distro switches will all have Cisco MACs attached.

r0berth1 - take a look at this UnDP for 6500 backplane

Np problem - Here is a thread regarding Avaya support . Are you using the 12000 on the inside of an MPLS core? If so, do you use different model access routers?

See the attached

Sorry for the acronym - PCAP is a packet capture. If you load Wireshark and capture packets on the NTA server we can see what the sources are sending. That will tell us if the issue is with the source exporter or NTA. Wireshark is freeware.

Please open a ticket so we can track the issue and fix. Andy

Can you post the trigger conditions?

DELETE FROM VoIPengines where servername='collector' From SQL Studio Express. Replace collector with the collector name you want to zap but leave the single quotes

If you are out of maintenance you can call customer service to get a license reset.

Hmmm. I can't duplicate that behavior. I'd try a browser cache clear and if that does not work put a call into support.

You need to examine the interface configs on the device. Look for IPSec and/or VPN configuration.

One method would be to create 2 groups. One with the PE, WAN router and FW and the other group with the switches and all other LAN side managed devices. Make the lan devices group dependent on the PE, router, FW group and make sure the status rollup on the PE, router, FW is set to worst

We do alert you in the Events resource when there is a flow that is being discarded. We don't interrogate the data that we discard so we can't give you any details on it.

I'll mark this for the Product Manager to look at.

The above chart is for actual TX and RX but will have a tend line. My system is on a lab network so traffic is basically zero.