High Memory Utilization on Orion Server

Go back to the basics:

1. Firstly open task manager and check which service is consuming the max memory. You need to start from there.

2. If its SolarWinds services - what service - each service has a different reason, based on what you see you can further dig into it accordingly.

3. I dont have enough context to suggest anything further - say for example if you are talking about your main polling engine and using the same to access your Web UI and Memory consumption is huge because of your IIS its a different story - could be the reason is max number of simultaneous users accessing the GUI or some user running a never ending report.

4. Is the issue with Collector service ?

5. Is the issue with your Job Engine service ?

6. Is it with Trap service ?

7. Is it with Syslog service ?

8. Is it with Netflow Service ?

........

Based on which service is consuming you can start digging in more as to what's happening and why is that service consuming that additional memory.

There are a lot of great replies out here and yes your windows firewall could be issue and you may need to allow a pass for SolarWinds services or it could be the anti virus, or it could be your regular maintenance or because of some windows patch or .net or etc etc etc

Hope this helps.

High memory usage in SolarWinds (especially nearing 99% with 64GB RAM) with NCM and NPM can be a complex issue. It often stems from a combination of factors, including configuration, resource demands, and potentially bugs within SolarWinds itself. While Tac may have not pinpointed the root cause, there are several areas to investigate and potential fixes. Cuyahoga County Auditor

absolutely, which is why i mentioned not sure whats eating up the memory and we wont know unless we have more context.

My Information service v3 touches 40-50GB and completely hangs the server.

Info service - ok in that case probably the same thing:

Get rid of IIS load if any try getting a AWS in place and dont use Main polling engine to host the Web UI

Get rid of load on the main poller try moving elements onto additional poller as much as you can

Windows patching and regular reboots (atleast once in a month to be in place)

Check the latency between your pollers and database - sometimes this as well plays a role 

And syslog or trap count should be well within the limits if not this might hamper your DB, if DB is at stake everything is including info service, as well need to make sure DB is healthy

And since NCM is also in place its good to check jobs, inventory runs, not communicating devices (connection errors) etc because of which i once had an issue with sub/pub clearing them helped me

Orion Platform Troubleshooting Guide - Services - SWIS

The above points should be a good starting point.

Info service - ok in that case probably the same thing:

Get rid of IIS load if any try getting a AWS in place and dont use Main polling engine to host the Web UI

Get rid of load on the main poller try moving elements onto additional poller as much as you can

Windows patching and regular reboots (atleast once in a month to be in place)

Check the latency between your pollers and database - sometimes this as well plays a role 

And syslog or trap count should be well within the limits if not this might hamper your DB, if DB is at stake everything is including info service, as well need to make sure DB is healthy

And since NCM is also in place its good to check jobs, inventory runs, not communicating devices (connection errors) etc because of which i once had an issue with sub/pub clearing them helped me

Orion Platform Troubleshooting Guide - Services - SWIS

The above points should be a good starting point.