CheckMK Monitoring - Number of threads

[Gustavo]

Hi,

The other day i see an alert coming from one Linux(RetHat) server, this alert was related to the Number of threads and the original threshold on the application:

more than 4k to critical state.

Something that I confirm on the server with this command:

ps -eo nlwp | tail -n +2 | awk ‘{ num_threads += $1 } END { print num_threads }’

But to bee hones, i would like to know, what kind of parameters do you, cmk administrator or sysadmin linux administrators, consider or use to validate if this is a real problem or nor, in my case, this number of threads, don´t generated affectations on the performance of the server.

What can i do to know if this is a real problem on the server, what do i need to see or know to decide?

Thank you.

[r.sander (Robert Sander)]

@Gustavo

What can i do to know if this is a real problem on the server, what do i need to see or know to decide?

It is totally up to you to change the check parameters to fit the situation on this system.
The thresholds are just default values.

[foobar]

we have servers where 20k is idle…
So find a proper value for your landscape and then set “exceptions” for the rest

[martin.schwarz]

We have configured relative levels for WARN/CRIT at 80%/90%
and have disabled absolute levels.

[randallpittman (Randy Pittman)]

I found that I needed to not only set relative levels to 80/90%, but I also needed to check the box for “Absolute levels” and then select “No Levels”.