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

SAM Polling Engine

Are there recommendations for the quantity of SAM components that the polling engine can handle? This polling engine has 955 assigned nodes and 9098 components. We are seeing a lot of CounterFetcher.exe and conhost.exe processes that are not terminating as expected. It seems that as these processes hang - the polling engine starts to have issues with SAM polling. We start to see a lot of components with the error "connection timeout: Job cancelled by scheduler" I am beginning to wonder if we have just overloaded the polling engine. We have a support case open, but no resolution yet.

10-11-2012 1-26-21 PM.png

Find more posts tagged with

Accepted answers

All comments

aLTeReGo

This screenshot shows only information regarding NPM/Core element counts (nodes/interfaces/volumes). I can say you're definitely nearing the higher end of a single poller for these elements. However, the screenshot shows nothing about SAM component monitors assigned to this polling engine, only that your at 66% of the maximum rate. How many component monitors are assigned to this additional poller?

klinejw

There are 9098 active SAM components on this polling engine.

bwicks

What version of SAM are you running. The upgrade to 5.2 fixed our issue with the hung CounterFetcher and ConHosts. . We have almost 11k components on 1 poller

klinejw

We are running SAM version 5.2. Actually its 5.2 RC3 - but we were told that there was no difference between RC3 and the GA release of 5.2.

bwicks - are you running any other Orion pollers on your poller with 11k components? We are running NPM, SAM, and SEUM pollers on the one we are having issues with.

bwicks

Yes we are running SEUM, SAM, UDT and IPSLA. How are you catching your components with "connection timeout: Job cancelled by scheduler" errors? do you have an alert setup or running a query to find them

klinejw

I wrote a query that allow me to filter on that Error Message. BTW - how many nodes do you have on that polling engine?

I have 40+ hung CounterFetcher and about the same number of ConHost processes on my polling engine at all times. It seems to take less than 24 hours after a complete restart for them to re-appear.

bwicks

1205 .. mix of servers and network gear. What server OS are you running. We are running 2008r2 64bit

klinejw

We are running 2008R2 as well. Our polling engine is managing almost all Windows servers and the majority of our deployed components are Performance Counter Monitor and WMI Monitor types. This polling engine is not short on hardware either - it has 24 cores and 16GB of RAM, but because of the hung processes it averages anywhere from 30 - 70% CPU, before the hung processes it runs around 10 - 20% CPU.

bwicks

Are the servers being monitored close network wise to the poller.. or are the servers spread around the country. I am wondering if it’s a timing issue. 90% of the servers we are monitoring are all in the same building as the SAM poller.