NSClient CPULoad: Division by zero

[Ralf Mellis]

Dean Bishop wrote:
> Just to clear my muddy brain...where and when exactly are you seeing this
> division by zero error and what exactly do you mean when you say that you
> are issuing the command "check_nt -v CPULOAD" on the server (i.e. on which
> server are you issuing the command).
> 

I have a nagios server with nagios 1.0 and nagios-plugins 1.3 (linux).
On all windows servers there is the nsclient 1.0.7.1 up and running.

I have defined several services for my hostgroups containing all windows 
servers. These services are launching the plugin "check_nt" (installed 
on the nagios server during plugin installation) with the options "-v 
CPULOAD", "-v USEDDISKSPACE" and "-v UPTIME" (with another parameters, 
as necessary). All windows servers are reporting fine as USEDDISKSPACE 
is concerned, three of my w2k servers are reporting "NSClient - 
ERROR:Malformed request or internal error. Check EventLog:None&2&10 " 
(this in the web gui of nagios) and "Division by zero " (this in the 
application log with source nsclient on the local w2k server) as CPULOAD 
is concerned. The UPTIME service for these three servers also does not 
report the correct values to the nagios server.
All other servers are reporting fine.
In addition I have checked the three windows servers locally with the 
sysinternals tool psinfo, on two of them this tool reports the correct 
values, on one of them it complains about "Error reading uptime".
So, IMHO it is obvious that regarding the UPTIME service the psinfo and 
nsclient service are _not_ querying the same source...

It is remarkable, that all my servers are on the same patch level as 
servicepacks are concerned (w2k --> sp3, nt4 --> sp6a and srp).
IMHO the given behaviour can't rely only on the servicepacks, it must 
have something to do with ie version and/or other installed hotfixes.

Now it is up to us, the ms users, to find out the reasons for the 
errors. I will try do this in the next time...

> 
> dean
> -----Original Message-----
> From: Ralf Mellis [mailto:testrm at kisters.com] 
> Sent: April 29, 2003 9:34 AM
> To: nagios-users at lists.sourceforge.net
> Subject: Re: [Nagios-users] NSClient CPULoad: Division by zero
> 
> Hello again,
> 
> now I have configured the "check_nt -v UPTIME" command on the three w2k 
> servers with the division by zero error issuing the "check_nt -v 
> CPULOAD" command.
> On all three servers nsclient reports "System Uptime : 0 day(s) 0 
> hour(s) 0 minute(s) " (!?).
> 
> As I mentioned before, two of theses three servers are reporting the 
> correct uptime using the sysinternals tool psinfo, one of them not.
> 
> Has anyone an idea, to correct the behaviour of theses servers 
> (registry, performance counters ore something)??
> 
> Bye
> 


-- 
Ralf Mellis
Abt. DV/ORG
Kisters Maschinenbau GmbH
Germany
47533 Kleve
Boschstr. 1-3
Tel. +49(0)2821-503-0
Fax  +49(0)2821-26110




-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
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