check_openmanage: timeout vs. SNMP timeout

[Trond Hasle Amundsen]

Andrew Daugherity <adaugherity at tamu.edu> writes:

> I'm troubleshooting an issue where one server is occasionally not responding (I think it's a firewall or snmpd issue, not this plugin), and I noticed that changing the timeout option to check_openmanage did not affect how long it took before receiving the
>   SNMP CRITICAL: No response from remote host A.B.C.D
>
> message.  Looking at the code I see the timeout option is _not_ passed to the Net::SNMP session object, so the SNMP connection timeout uses the default value (5 seconds according to the Net::SNMP man page, but 10 seconds in my testing).
>
> If I pass the timeout option to the Net::SNMP->session object like so:
> ====
> diff --git a/check_openmanage b/check_openmanage
> index b6abec5..3558ed4 100755
> --- a/check_openmanage
> +++ b/check_openmanage
> @@ -860,6 +860,7 @@ sub snmp_initialize {
>          '-port'     => $opt{port},
>          '-hostname' => $opt{hostname},
>          '-version'  => $opt{protocol},
> +        '-timeout'  => $opt{timeout},
>         );
>  
>      # Setting the domain (IP version and transport protocol)
> ====
> Then it does obey the timeout option and I instead get the
>   PLUGIN TIMEOUT: check_openmanage timed out after 30 seconds
>
> message.  This might be by design though, to have a shorter SNMP timeout and different error messages, but it was perplexing to me why the timeout option was seemingly not working.  Perhaps a different option for the SNMP timeout, or a documentation clarification, is a better way?

Hello Andrew,

Your analysis of this problem is correct, you're hitting the Net::SNMP
timeout which is default 5 seconds. There are two reasons why the
--timeout parameter isn't passed to the SNMP object:

  1. I never saw any reason to :) This is the first time I've heard of
     problems relating to it.

  2. The SNMP object timeout has limitations, it can only be between 1
     and 60 seconds. I don't know how Net::SNMP reacts if the specified
     value is outside of this range.

The documentation is lacking on this, as you pointed out, and I'll fix
that. A new option to specify the SNMP object timeout would be easy to
add, and is in my opinion a cleaner approach than just passing the
plugin timeout.

PS. I'm going away for the weekend and I'm leaving in a few minutes, so
I'll get back to you on this early next week.

Regards,
-- 
Trond H. Amundsen <t.h.amundsen at usit.uio.no>
Center for Information Technology Services, University of Oslo

------------------------------------------------------------------------------
LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial
Remotely access PCs and mobile devices and provide instant support
Improve your efficiency, and focus on delivering more value-add services
Discover what IT Professionals Know. Rescue delivers
http://p.sf.net/sfu/logmein_12329d2d
_______________________________________________
Nagios-users mailing list
Nagios-users at lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-users
::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. 
::: Messages without supporting info will risk being sent to /dev/null