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

Poller Hang Detection?

I need a way to detect when my polling system hangs as this is happening relatively regularly and seems due to database connection problems.

One of the things I have noticed when this happens is that my SNMP Outstanding statistic on the Polling Engines page is very high, above 2000. Is there some place in the database where this is stored so I can monitor this statistic as an early indicator of problems?

Thanks in advance for any help with this problem!

Find more posts tagged with

Accepted answers

All comments

njoylif

It sounds like your polling engine is overloaded. I was told when troubleshooting my similar issues that if the outstanding gets above 900, the engine will "go to lunch" and not come back.

that being said, the engines table has the info you want...
Below is a SQL query that will give you the output you are looking for...notice MMORIV02 is "out to lunch" and the polling completion shot up to 99.9xxx% -

select ServerName, IP, KeepAlive, Elements, Nodes, Interfaces, pollers, pollingcompletion from engines

robertcbrowning

We suffered from this too. 1 primary + 4 pollers with all the modules and a web site too. Every so often a polling engine would just stop. Sometimes it would still poll, but report that it was not. Quite confusing. We had NCM running on the primary and it occasionally it took up 100% of 2 cores and most of the available memory, leaving a couple of cores and not much RAM to hold everything else together. We moved the NCM to a new server and it's been rock-solid ever since. We've always tried to keep the primary faily empty of elements (less than 4000) to help with the loading, but we'll be moving some more away as soon as we get more pollers.

byrona

Well, the polling engine is heavily loaded but I don't believe it's overloaded because it normally functions just fine. There is some event that is causing it to get behind and that is what I am trying to isolate and it looks to be a database connection issue in addition to Syslog service eating up most of the CPU.

I looked in the Engines table but I don't see the SNMP Outstanding statistic that I am looking for.

njoylif

gotcha. I have no idea where (what table or calc) they get that number from. I think it has been on thwack that people want web based access to those stats as that is really good info...