Community
- Command Central
- MVP Program
- Monthly Mission
- Blogs
- Groups
- Events
- Media Vault
Products
- Observability
- Network Management
- Application Management
- IT Security
- IT Service Management
- System Management
- Database Management
Content Exchange
- SolarWinds Platform
- Server & Application Monitor
- Database Performance Analyzer
- Server Configuration Monitor
- Network Performance Monitor
- Network Configuration Manager
- SQL Sentry
- Web Help Desk
Free Tools & Trials
Store

Reducing/deleting nodes or interfaces from existing polling engines

Thwack Forum,

I am working at a client site where there are multiple pollers installed. While running Network Sonar discoveries, we have

inadvertently overloaded 2 of 6 pollers.

How can someone eliminate specific network nodes or interfaces from an overloaded poller?

Thanks in advance.

kmnruser

Find more posts tagged with

Accepted answers

jamison.jennings

KMNRuser Just a couple of notes here that could possibly help you out.

Manage Nodes screen gives you a ton of flexibility here.

First, if you have overloaded pollers, you could Group by Polling Engine in manage nodes then click on each polling engine to see nodes, and if you go to the top and toggle Nodes to interfaces, you could quickly see what Nodes/Interfaces are on each poller.

You could also select the check box next to a node or several nodes and under the "More Actions" box, quickly move nodes to underloaded pollers by selecting "Change Polling Engine" to balance out your load.

Second, if you are truly starting from scratch with your deployment, I would create a new database with the re-install of NPM. That way there is no carry over from old DB tables and you would truly be clean.

I would get an export of nodes your monitoring now for easier re-discovery.

All comments

mrxinu

You could use the API to delete the nodes and interfaces you don't want if you can come up with a method for identifying them (e.g., all loopback interfaces). If you're interested in that I'll send a pull request to the SolarWinds OrionSDK repository for examples of both.

Steven W. Klassen

Programmer Analyst @ Loop1 Systems

http://www.loop1systems.com/

http://www.linkedin.com/in/mrxinu

KMNRuser

Steven,

We are actually getting ready to start from scratch again with our deployment. The goal in a perfect world would be a clean database.

We were thinking that by deleting all nodes it would clean the database. Perhaps you can comment on if that will work.

I would think that deleting all nodes would implicitly delete all associated interfaces with that node.

Is this correct?

And to answer your question, please do post the pull request so we can look at it.

KMNRuser

Steven,

Also could you please explain what you mean specifically when you say "use the API"?..

Thanks,

KMNRuser

jamison.jennings

KMNRuser Just a couple of notes here that could possibly help you out.

Manage Nodes screen gives you a ton of flexibility here.

I would get an export of nodes your monitoring now for easier re-discovery.

mrxinu

I agree completely with jamison.jennings on the fresh database. You could delete all the nodes which would take the interfaces with it but it wouldn't truly be a clean slate if you're concerned about cruft hanging around from previous upgrades. If you just want an empty inventory though, that'll do the trick.

Steven W. Klassen

Programmer Analyst @ Loop1 Systems

http://www.loop1systems.com/

http://www.linkedin.com/in/mrxinu

mrxinu

Hi KMNRuser! There's a development toolkit from SolarWinds that gives you access to PowerShell cmdlets that let you query the SolarWinds inventory for details about the nodes and interfaces and then you can use it to delete any interfaces that you don't want any longer.

Here's a link to the SDK and some PowerShell examples on SolarWinds' GitHub repository:

Releases · solarwinds/OrionSDK · GitHub

OrionSDK/Samples/PowerShell at master · solarwinds/OrionSDK · GitHub

Steven W. Klassen

Programmer Analyst @ Loop1 Systems

http://www.loop1systems.com/

http://www.linkedin.com/in/mrxinu

cgregors

This has been a common problem I've dealt with. We have 3 instances of Orion and one has 6 pollers. When one gets overloaded, we shift nodes to other pollers to balance the load. The only issue with load balancing may be firewall rules that are in place for specific pollers. If there are firewalls in the way, it's good to have multiple pollers on the far side of the firewalls so you can load balance between them.

Related to discovery, Orion will by default assign the nodes to the poller that was used to run the discovery. In that case, I've found that load balancing the discoveries between pollers is the answer.

As to moving nodes between pollers, go to manage nodes, group by polling engine, select a bunch and multi-edit them to change the polling engine. You may still end up with a conflict between the engine you're using for polling and the engine that the discovery runs from.

It's a pain.

Chris