Difference between Service check timeout and Agent TCP timeout?

[davidwayne]

Can anybody explain me the difference between the two ?

[openmindz (Thomas Karatzoglou)]

Hi @davidwayne

I assume that you’re following up yourself on your other thread: Keepalive time out message

At the risk of “stating the obvious”: The “online help” for each rule mostly can explain what a rule is for. You can reach this by clicking in the little “book” icon on the top right corner. That said, the explanations I can find for those two particular rules, are the following ones:

• Service check timeout (Microcore): Using this option you can override the globally configured setting for check timeouts for individual services.

• Agent TCP connect timeout: Timeout for TCP connect to the Check_MK agent in seconds. If the connection to the agent cannot be established within this time, it is considered to be unreachable. Note: This does not limit the time the agent needs to generate its output. This rule can be used to specify a timeout on a per-host-basis.

HTH,
Thomas

[davidwayne]

Thanks Thomas.
I am not sure if my problem with the keepalive timeout will be resolved with either of these options. However, AFAIU both of these rules point to Check_MK service check (which is the only active check). I have few Linux servers on which a nightly backup runs and this causes the keepalive timeout. Now, which option is best suited for my scenario ?

[openmindz (Thomas Karatzoglou)]

Hey @davidwayne

The “Service check timeout” allows to (and I quote again): “override the globally configured setting for check timeouts for individual services”

The other rule is only modifying the connection timeout to the CMK agent, that’s correct. Probably the second option should do the trick. Be bold… Experiment…

Thomas

[davidwayne]

The service check timeout gives the desired result atleast on the first glance.