NSClient++ Help (#1) - Errors in nsclient.log (#1028) - Message List
Need a little help with NSClient++ running with Nagios Core
Sometimes my servers stop responding and I get the following message in Nagios:
CRITICAL - Socket timeout after 10 seconds
which seems to correspond to the below error in nsclient.log on the client machine:
error:modules\NSCAAgent\NSCAThread.cpp:312: Timeout reading NSCA hdr packet (increase socket_timeout), we only got: 0
This particular machine is running server 2003 but it seems to happen on a couple of different machines.
Not much on the web regarding the error so looking for any advice??
I should probably mention I'm using Nagios using only passive checks using NSCA and my Nagios is the core version running on CentOS 5.8
I've had some issues with the NSCA process on that side of things but this seems to be a NSClient side errorjer198208/09/12 14:37:31 (10 months ago)
Well, message leads me to believe you have timeout issues between the client and server. To be specific NSCA client (NSClient++) connects to NSCA server and cant get the initial encryption vector data in the allotted time.
If you in fact have network issues I would start by resolving them :) I think you are on 0.3.9 there?
In 0.4.0 the NSCA handling has been improved meaning a lot of errors are covered better and overall improvements. So by upgrading things could magically become better. In 0.4.1 I am currently adding retry so you could then "resolve" the problem by retrying... in 0.4.2 (maybe 0.4.1) caching will be implemented as well meaning you will be able to "resubmit once the network is ok" again).
But as I said before if you have network issues maybe easier to resolve them and if you dont have network issues it could be a bug (but then I would need something a bit more specific network dumps maybe, or debug logs?)
Michael Medinmickem08/09/12 14:56:04 (10 months ago)
Thanks for the reply Michael - I'm hoping we don't have network issues, everything is showing as OK in Cacti and nothing else is showing a problem although our clients are outside our network behind their own routers and firewalls (hence the reason I'm using passive checks only).
I'll try a upgrade to 0.4.1 on one of the problem clients and see how that goes.
Will report back soon, thanks!jer198208/09/12 16:14:35 (10 months ago)