NSClient++ log growing, error repeating

[jpipitone]

UPDATE:

This is happening on ALL instances of Windows installations of the NSClient++.

This error is repeating itself over and over, causing the log file on a particular machine to grow out of control.

I've tried searching only to find that I need to upgrade the NSclient to solve the issue, however after a re-install, it still fills up the log file.

Any ideas what could be causing this?

2011-06-21 14:55:57: error:modules\CheckSystem\PDHCollector.cpp:215: Failed to query performance counters: \Processor(_total)\% Processor Time: PdhGetFormattedCounterValue failed: A counter with a negative denominator value was detected. (800007D6)

Server: Windows 2003 Enterprise Edition 32 bit
NSClient++ version: NSClient++ - 0.3.8.75

[mguthrie]

Can you show us the check command and service definition for that particular service?

[jpipitone]

Check command:

$USER1$/check_nt -H $HOSTADDRESS$ -p 12489 -v CPULOAD -l 5,90,95,10

Service Definition:

I'm not sure where to find this as this was imported from a Nagios 2.9 core installation. I can't find the service definition anywhere.

Could this be as simple as replacing the legacy Nagios 2.9 checks with NagiosXI checks?

[jpipitone]

One more thing to note - I'm seeing this error on Windows machines where we're not even checking CPU. It's also happening multiple times per second.

For instance, on a particular server, we are only checking the disk usage, and I still see this in the NSClient log from today:

2011-06-21 09:50:00: error:modules\CheckSystem\PDHCollector.cpp:215: Failed to query performance counters: \Processor(_total)\% Processor Time: PdhGetFormattedCounterValue failed: A counter with a negative denominator value was detected. (800007D6)

[mguthrie]

Is this something that just recently started at the same time for all windows servers, or is it something that's been ongoing for quite some time?

[jpipitone]

It looks like it has been going on since we purchased and installed NagiosXI - Nagios sent us a notification that a C drive had filled up, and I used a utility to find the largest files on the drive, the NSCLient.log being one of them - around 4gb in size.

Yesterday, I just checked our other Windows servers and this is also the case with the log files.

I've also noticed that even if I replace the legacy check with a new CPU load check using the Nagios Wizard, the errors still appear and are being logged.

[mguthrie]

2011-06-21 09:50:00: error:modules\CheckSystem\PDHCollector.cpp:215:

This is error output from one of the NSClient++ modules, have you come across any other information on this from http://nsclient.org/nscp/

[jpipitone]

Dug around a bit, apparently it's a bug.

I've commented out filelogger.dll to stop the logging completely until the problem gets patched. Thanks for your help either way.