I’ve been using Nagios for a little while now and have recently configured Nagios to use NSCP (passive service checks).
I have a distributed setup and this works extremely well.
What I would love to know is how to setup passive host checks. I have read the relevant documentation, however, it does not explicitly detail how to configure NSClient++ to send such a check to my nagios server. I am still learning to use Nagios and would like some advice.
What exactly needs to be setup on the server/client ends? Anyone have any examples they are willing to share?
Wht kind of checks are you doing, how many servers?
Where is the active part located and where the passive. Using NSCP you are not necessarily doing passive checks. (that would be NRPE or NSCA).
My apologies, luca, I am infact using NSCA (long day at the office yesterday).
All checks (CPU, Memory, Windows Servcies, Disk Capacity, Uptime, etc) are done passively (NSClient++ reports service status to my Nagios server via NSCA).
I have one check_rbl (spam blacklist) service which is actively checked by Nagios.
Each client will have approximately 7 service checks each.
I have check_stale in place to notify me (service status turns critical) when a client fails to report the status of a service after so many minutes, I would like to do something similar with the host check.
Host checks are currently being actively checked by Nagios, but this is not ideal as each server sits behind its own firewall.
This means active checks will only tell me whether the customer router is online or not, which I am not really interested in at this stage.
NAT is not a valid option as multiple servers sit behind one firewall and I want to keep the configuration as simple as possible.
When the rollout is complete, we will be monitoring between 80-100 Windows servers.
Back to initial question then
I think passive host checks don’t make much sense, each time a service is checked you already have a passive host check, if the service is alive the host HAS to be alive. The fact that something gets to the nagios server is in itself a confirmation that the host is alive. When would you need to check the host then? if NSCP fails… but then you wouldn’t be able to make a passive check anymore. I think you only need active host checks from the nagios server (they usually get done only if a service fails, in case of passive checks that would be a stale check if i rmeber right) OR you make the host checks form a third server, in which case they would in fact be passive, but that doens’t look like an option…
Hope this helps, or did i miss a step?
EDIT: You could in fact create a small script which makes the passive checks somehow… nagios.sourceforge.net/docs/3_0/ … hecks.html
Have the remote host “ping” the server with a small script… , log it and check the log file, and write something in the nagios command file accordingly to the doc… might work.(easisest would be a telnet on port 80 and using the apache log probably). I’m just not sure if it’s worth the effort…
