Why is MRTG getting values of zero for our loadbalancer

Why is MRTG getting values of zero for our loadbalancer

Post by Malcolm Ferguso » Sun, 06 Jul 2003 01:37:11



I've tried to configure MRTG to monitor traffic and resource utilisation
on our Radware loadbalancer.  Unfortunately, MRTG is plotting zeroes for
everything.

For the interfaces, every entry in the log file is zero.  For the
resource utilisation, the first entry is correct (confirmed with
snmputil from the Windows Resource Kit and snmpwalk from Linux), then
everything else is zero.  I can poll the OID from the command line and
get the value I'm expecting.  The numbers I see with these utils
correspond to the graph that I have Radware's ConfigWare is plotting.
But MRTG is only plotting zeroes.

Platform: Windows 2000 Terminal Server
Perl: Active State's ActivePerl 5.6.1 Build 635
MRTG: 2.9.29

Cheers,
Malc

 
 
 

Why is MRTG getting values of zero for our loadbalancer

Post by Malcolm Ferguso » Sun, 06 Jul 2003 02:31:10



> I've tried to configure MRTG to monitor traffic and resource utilisation
> on our Radware loadbalancer.  Unfortunately, MRTG is plotting zeroes for
> everything.

> For the interfaces, every entry in the log file is zero.  For the
> resource utilisation, the first entry is correct (confirmed with
> snmputil from the Windows Resource Kit and snmpwalk from Linux), then
> everything else is zero.  I can poll the OID from the command line and
> get the value I'm expecting.  The numbers I see with these utils
> correspond to the graph that I have Radware's ConfigWare is plotting.
> But MRTG is only plotting zeroes.

> Platform: Windows 2000 Terminal Server
> Perl: Active State's ActivePerl 5.6.1 Build 635
> MRTG: 2.9.29

> Cheers,
> Malc

It seems to be keeping an empty log every time it writes.  For instance,
this is the top of the log from two time instances:

First:
1057340871 12 12
1057340871 0 0 0 0
1057340571 0 0 0 0
1057340400 0 0 0 0
1057340100 0 0 0 0

Later:
1057342971 13 13
1057342971 0 0 0 0
1057342671 0 0 0 0
1057342500 0 0 0 0
1057342200 0 0 0 0
1057341900 0 0 0 0
1057341600 0 0 0 0
1057341300 0 0 0 0
1057341000 0 0 0 0
1057340700 0 0 0 0
1057340400 0 0 0 0
1057340100 0 0 0 0

 
 
 

Why is MRTG getting values of zero for our loadbalancer

Post by Joachim Pens » Sun, 06 Jul 2003 03:20:35




>>I've tried to configure MRTG to monitor traffic and resource utilisation
>>on our Radware loadbalancer.  Unfortunately, MRTG is plotting zeroes for
>>everything.

>>For the interfaces, every entry in the log file is zero.  For the
>>resource utilisation, the first entry is correct (confirmed with
>>snmputil from the Windows Resource Kit and snmpwalk from Linux), then
>>everything else is zero.  I can poll the OID from the command line and
>>get the value I'm expecting.  The numbers I see with these utils
>>correspond to the graph that I have Radware's ConfigWare is plotting.
>>But MRTG is only plotting zeroes.

>>Platform: Windows 2000 Terminal Server
>>Perl: Active State's ActivePerl 5.6.1 Build 635
>>MRTG: 2.9.29

>>Cheers,
>>Malc

> It seems to be keeping an empty log every time it writes.  For instance,
> this is the top of the log from two time instances:

> First:
> 1057340871 12 12

So you had cumulatively 12

Quote:> 1057340871 0 0 0 0
> 1057340571 0 0 0 0
> 1057340400 0 0 0 0
> 1057340100 0 0 0 0

and this value did not increase, hence correctly no traffic

Quote:

> Later:
> 1057342971 13 13
> 1057342971 0 0 0 0
> 1057342671 0 0 0 0
> 1057342500 0 0 0 0
> 1057342200 0 0 0 0
> 1057341900 0 0 0 0
> 1057341600 0 0 0 0
> 1057341300 0 0 0 0
> 1057341000 0 0 0 0
> 1057340700 0 0 0 0
> 1057340400 0 0 0 0
> 1057340100 0 0 0 0

In many cycles, the cumulative value rose to 13, but this increase by
one is rounded away. You just have very little traffic.

Or, isn't it in/out-octets you measure? In that case, make sure that you
specify "gauge" in the config file, than you will not get the
differences of subsequent values but the actual values stored and plotted.

Joachim

 
 
 

Why is MRTG getting values of zero for our loadbalancer

Post by Malcolm Ferguso » Sun, 06 Jul 2003 03:45:26



<snip>
> In many cycles, the cumulative value rose to 13, but this increase by
> one is rounded away. You just have very little traffic.

> Or, isn't it in/out-octets you measure? In that case, make sure that you
> specify "gauge" in the config file, than you will not get the
> differences of subsequent values but the actual values stored and plotted.

Sorry, I wasn't being clear.  That was from the Resource Utilisation
log.  It's a value between 0 and 120 according to one source, and a
percentage according to another.  This is the config for it:

Options[_]: growright, bits
WorkDir: C:\inetpub\wwwroot\MRTG
RunAsDaemon: yes

Target[radware1_cpu]:

MaxBytes[radware1_cpu]: 120
Directory[radware1_cpu]: <x.x.x.x_x>
#AbsMax[radware1_cpu]: 120
Title[radware1_cpu]: Loadbalancer: Resource Utilization
#Unscaled[radware1_cpu]: d
WithPeak[radware1_cpu]: ymw
YLegend[radware1_cpu]: Usage[0:120]
ShortLegend[radware1_cpu]: CPU
Legend1[radware1_cpu]: Average Resource Utilization
Legend2[radware1_cpu]: Average Resource Utilization
Legend3[radware1_cpu]: Maximal Resource Utilization
Legend4[radware1_cpu]: Maximal Resource Utilization
PageTop[radware1_cpu]: <H1>Loadbalancer Resource Utilization
 </H1>
 <TABLE>
   <TR><TD>System:</TD><TD>Loadbalancer</TD></TR>
   <TR><TD>Maintainer:</TD><TD>x</TD></TR>
  </TABLE>

 
 
 

Why is MRTG getting values of zero for our loadbalancer

Post by Joachim Pens » Sun, 06 Jul 2003 16:36:12




> <snip>

>>In many cycles, the cumulative value rose to 13, but this increase by
>>one is rounded away. You just have very little traffic.

>>Or, isn't it in/out-octets you measure? In that case, make sure that you
>>specify "gauge" in the config file, than you will not get the
>>differences of subsequent values but the actual values stored and plotted.

> Sorry, I wasn't being clear.  That was from the Resource Utilisation
> log.  It's a value between 0 and 120 according to one source, and a
> percentage according to another.  This is the config for it:

> Options[_]: growright, bits

As I said, add a ", gauge" to this line. From the MRTG doc:

|gauge
|    Treat the values gathered from target as 'current status'
|measurements and not as ever incrementing counters. This would be
|useful to monitor things like disk space, processor load, temperature,
|and the like ...

|    In the absence of 'gauge' or 'absolute' options, MRTG treats
|variables as a counters and calculates the difference between the
|current and the previous value and divides that by the elapsed time
|between the last two readings to get the value to be plotted.

Joachim

 
 
 

Why is MRTG getting values of zero for our loadbalancer

Post by Malcolm Ferguso » Wed, 09 Jul 2003 23:21:17





> > <snip>

> >>In many cycles, the cumulative value rose to 13, but this increase by
> >>one is rounded away. You just have very little traffic.

> >>Or, isn't it in/out-octets you measure? In that case, make sure that you
> >>specify "gauge" in the config file, than you will not get the
> >>differences of subsequent values but the actual values stored and plotted.

> > Sorry, I wasn't being clear.  That was from the Resource Utilisation
> > log.  It's a value between 0 and 120 according to one source, and a
> > percentage according to another.  This is the config for it:

> > Options[_]: growright, bits

> As I said, add a ", gauge" to this line. From the MRTG doc:

Great!  That fixed it.  I didn't take in your comment about the "gauge"
option as the paragraph appeared to be about network throughput, not CPU
load.  My mistake.

Cheers,
Malc

x--   100 Proof News - http://www.100ProofNews.com
x--   3,500+ Binary NewsGroups, and over 90,000 other groups
x--   Access to over 800 Gigs/Day - $8.95/Month
x--   UNLIMITED DOWNLOAD

 
 
 

1. getting CPU usage value for MRTG

I'm trying to monitor CPU usage via MRTG on a box running UCD-SNMPD. I've
got it working using the OID laLoadInt ( 1.3.6.1.4.1.2021.10.1.5.1 - CPU
load avereges). But these values are a lot higher than the values I get when
I run TOP, or when I get values from ssCpuSystem.0 (percentage of system CPU
time), ssCpuUser.0 (percentage of user CPU time) and ssCpuIdle.0 (percentage
of idle CPU time).
Can anyone tell me what are the difference between these values?
Wich one is the most indicated do monitor CPU usage?

Thanks,
Claudio

2. HP IIIP -- sudden death: power supply ?

3. Why am I getting junk mail on my cable ISP???

4. ruling paper with thin lines

5. 'lame server resolving' - why am i getting this?

6. Executing Command During Notes Client Shutdown

7. ipRouteIfIndex value of zero?

8. removing plain password hack from Win95 registry

9. Zero values

10. MRTG-graph of week/month/year from some interfaces show zero

11. MRTG Outgoing always zero

12. Just Dial One-Nine-Zero-Zero-SCREWME

13. HELP! Why zero length files in print queue?