[mrtg] Target did not eval into defined data problem

[Attila Rajmund Nohl]

Hello!

I have the following lines in my mrtg.cfg:

Target[eselivm3v80l_memory]:
.1.3.6.1.4.1.193.105.1.1.1.21.1.0&.1.3.6.1.4.1.193.105.1.1.1.21.1.0:***@eselivm3v80l:4000::::2
Title[eselivm3v80l_memory]: Memory usage of NETSim
MaxBytes[eselivm3v80l_memory]: 10000000
RouterUptime[eselivm3v80l_memory]: ***@eselivm3v80l:4000
Options[eselivm3v80l_memory]: growright,nopercent,unknaszero
LegendI[eselivm3v80l_memory]: Memory usage of NETSim, as reported by the OS
LegendO[eselivm3v80l_memory]: Memory usage of NETSim, as reported by the OS
ShortLegend[eselivm3v80l_memory]: bytes
YLegend[eselivm3v80l_memory]: Bytes

I can get the necessary values via SNMP:

> snmpget -v2c -c public eselivm3v80l:4000 .1.3.6.1.4.1.193.105.1.1.1.21.1.0
NME-MIB::netsimVMTotalMemoryUsageByOS.0 = Counter64: 339832832 bytes
> snmpget -v2c -c public eselivm3v80l:4000 sysUpTime.0
DISMAN-EVENT-MIB::sysUpTimeInstance = Timeticks: (333907) 0:55:39.07

Still I get these errors from mrtg:

ERROR: Target[eselivm3v80l_memory][_IN_] ' $target->[34]{$mode} ' did
not eval into defined data
ERROR: Target[eselivm3v80l_memory][_OUT_] ' $target->[34]{$mode} ' did
not eval into defined data

Do you know why? I'm using mrtg-2.12.2 on SuSE 10 SP3.

[M***@jwpepper.com]

Hey Attila, just going out on a limb here, but it looks like you
"Target[]:" line should have 5 colons at the end of the line before the 2.

Target[eselivm3v80l_memory]:
.1.3.6.1.4.1.193.105.1.1.1.21.1.0&.1.3.6.1.4.1.193.105.1.1.1.21.1.0:***@eselivm3v80l:4000:::::2

Let me know if that helps...

-Matt





From: Attila Rajmund Nohl <***@gmail.com>
To: <***@lists.oetiker.ch>
Date: 06/07/2012 01:42 PM
Subject: [mrtg] Target did not eval into defined data problem
Sent by: <mrtg-bounces+mmartin=***@lists.oetiker.ch>



Hello!

I have the following lines in my mrtg.cfg:

Target[eselivm3v80l_memory]:
.1.3.6.1.4.1.193.105.1.1.1.21.1.0&.1.3.6.1.4.1.193.105.1.1.1.21.1.0:***@eselivm3v80l:4000::::2
Title[eselivm3v80l_memory]: Memory usage of NETSim
MaxBytes[eselivm3v80l_memory]: 10000000
RouterUptime[eselivm3v80l_memory]: ***@eselivm3v80l:4000
Options[eselivm3v80l_memory]: growright,nopercent,unknaszero
LegendI[eselivm3v80l_memory]: Memory usage of NETSim, as reported by the
OS
LegendO[eselivm3v80l_memory]: Memory usage of NETSim, as reported by the
OS
ShortLegend[eselivm3v80l_memory]: bytes
YLegend[eselivm3v80l_memory]: Bytes

I can get the necessary values via SNMP:

> snmpget -v2c -c public eselivm3v80l:4000
.1.3.6.1.4.1.193.105.1.1.1.21.1.0
NME-MIB::netsimVMTotalMemoryUsageByOS.0 = Counter64: 339832832 bytes
> snmpget -v2c -c public eselivm3v80l:4000 sysUpTime.0
DISMAN-EVENT-MIB::sysUpTimeInstance = Timeticks: (333907) 0:55:39.07

Still I get these errors from mrtg:

ERROR: Target[eselivm3v80l_memory][_IN_] ' $target->[34]{$mode} ' did
not eval into defined data
ERROR: Target[eselivm3v80l_memory][_OUT_] ' $target->[34]{$mode} ' did
not eval into defined data

Do you know why? I'm using mrtg-2.12.2 on SuSE 10 SP3.

[Attila Rajmund Nohl]

Hello!

Actually I do get SNMP errors with five colons, because these
variables are at the end of the MIB and the SNMP GetBulk reaches the
end of MIB.

2012/6/7 <***@jwpepper.com>:
> Hey Attila, just going out on a limb here, but it looks like you "Target[]:"
> line should have 5 colons at the end of the line before the 2.
>
> Target[eselivm3v80l_memory]:
> .1.3.6.1.4.1.193.105.1.1.1.21.1.0&.1.3.6.1.4.1.193.105.1.1.1.21.1.0:***@eselivm3v80l:4000:::::2
>
> Let me know if that helps...
>
> -Matt
>
>
>
>
>
> From:        Attila Rajmund Nohl <***@gmail.com>
> To:        <***@lists.oetiker.ch>
> Date:        06/07/2012 01:42 PM
> Subject:        [mrtg] Target did not eval into defined data problem
> Sent by:        <mrtg-bounces+mmartin=***@lists.oetiker.ch>
> ________________________________
>
>
>
> Hello!
>
> I have the following lines in my mrtg.cfg:
>
> Target[eselivm3v80l_memory]:
> .1.3.6.1.4.1.193.105.1.1.1.21.1.0&.1.3.6.1.4.1.193.105.1.1.1.21.1.0:***@eselivm3v80l:4000::::2
> Title[eselivm3v80l_memory]: Memory usage of NETSim
> MaxBytes[eselivm3v80l_memory]: 10000000
> RouterUptime[eselivm3v80l_memory]: ***@eselivm3v80l:4000
> Options[eselivm3v80l_memory]: growright,nopercent,unknaszero
> LegendI[eselivm3v80l_memory]: Memory usage of NETSim, as reported by the OS
> LegendO[eselivm3v80l_memory]: Memory usage of NETSim, as reported by the OS
> ShortLegend[eselivm3v80l_memory]: bytes
> YLegend[eselivm3v80l_memory]: Bytes
>
> I can get the necessary values via SNMP:
>
>> snmpget -v2c -c public eselivm3v80l:4000 .1.3.6.1.4.1.193.105.1.1.1.21.1.0
> NME-MIB::netsimVMTotalMemoryUsageByOS.0 = Counter64: 339832832 bytes
>> snmpget -v2c -c public eselivm3v80l:4000 sysUpTime.0
> DISMAN-EVENT-MIB::sysUpTimeInstance = Timeticks: (333907) 0:55:39.07
>
> Still I get these errors from mrtg:
>
> ERROR: Target[eselivm3v80l_memory][_IN_] ' $target->[34]{$mode} ' did
> not eval into defined data
> ERROR: Target[eselivm3v80l_memory][_OUT_] ' $target->[34]{$mode} ' did
> not eval into defined data
>
> Do you know why? I'm using mrtg-2.12.2 on SuSE 10 SP3.
>
> _______________________________________________
> mrtg mailing list
> ***@lists.oetiker.ch
> https://lists.oetiker.ch/cgi-bin/listinfo/mrtg
>
>

[Steve Shipway]

The error message shows that MRTG could not get a valid result from the SNMP query for the OIDs specified.

Target[eselivm3v80l_memory]:
.1.3.6.1.4.1.193.105.1.1.1.21.1.0&.1.3.6.1.4.1.193.105.1.1.1.21.1.0:***@eselivm3v80l:4000::::2

Omit the leading '.' on the OIDs. Then all will be OK:

Target[eselivm3v80l_memory]:
1.3.6.1.4.1.193.105.1.1.1.21.1.0&1.3.6.1.4.1.193.105.1.1.1.21.1.0:***@eselivm3v80l:4000::::2

Steve


Steve Shipway
University of Auckland ITS
UNIX Systems Design Lead
***@auckland.ac.nz
Ph: [phone] ext 86487

[Attila Rajmund Nohl]

2012/6/7 Steve Shipway <***@auckland.ac.nz>:
> The error message shows that MRTG could not get a valid result from the SNMP query for the OIDs specified.
>
> Target[eselivm3v80l_memory]:
> .1.3.6.1.4.1.193.105.1.1.1.21.1.0&.1.3.6.1.4.1.193.105.1.1.1.21.1.0:***@eselivm3v80l:4000::::2
>
> Omit the leading '.' on the OIDs.  Then all will be OK:
>
> Target[eselivm3v80l_memory]:
> 1.3.6.1.4.1.193.105.1.1.1.21.1.0&1.3.6.1.4.1.193.105.1.1.1.21.1.0:***@eselivm3v80l:4000::::2

Now I don't get errors, but the data is not updated either. This is
how the first few lines of the <target>.log file look like:

1339164901 340201472 340201472
1339164901 0 0 0 0
1339164611 0 0 0 0
1339164600 0 0 0 0
1339164300 0 0 0 0

And 10 minutes later:

1339165501 530710528 530710528
1339165501 0 0 0 0
1339165211 0 0 0 0
1339165200 0 0 0 0
1339164900 0 0 0 0
1339164600 0 0 0 0
1339164300 0 0 0 0
1339164000 0 0 0 0
1339163700 0 0 0 0

[Steve Shipway]

Possibly you should be using the 'gauge' option? Is this variable a counter, or an absolute measure?

Also the rate of change may be < 1 per sec. If this is the case, native mode MRTG cannot cope with it, you need RRDTool as your backend.

Steve

Steve Shipway
University of Auckland ITS
UNIX Systems Design Lead
***@auckland.ac.nz
Ph: [phone] ext 86487


________________________________________

Now I don't get errors, but the data is not updated either. This is
how the first few lines of the <target>.log file look like:

1339164901 340201472 340201472
1339164901 0 0 0 0
1339164611 0 0 0 0
1339164600 0 0 0 0
1339164300 0 0 0 0

***@lists.oetiker.ch
https://lists.oetiker.ch/cgi-bin/listinfo/mrtg

[Attila Rajmund Nohl]

Hello!

This is an absolute measure. It's memory usage of a process and is
only updated every 15 minutes.

2012/6/9 Steve Shipway <***@auckland.ac.nz>:
> Possibly you should be using the 'gauge' option?  Is this variable a counter, or an absolute measure?
>
> Also the rate of change may be < 1 per sec.  If this is the case, native mode MRTG cannot cope with it, you need RRDTool as your backend.
>
> Steve
>
> Steve Shipway
> University of Auckland ITS
> UNIX Systems Design Lead
> ***@auckland.ac.nz
> Ph: [phone] ext 86487
>
>
> ________________________________________
>
> Now I don't get errors, but the data is not updated either. This is
> how the first few lines of the <target>.log file look like:
>
> 1339164901 340201472 340201472
> 1339164901 0 0 0 0
> 1339164611 0 0 0 0
> 1339164600 0 0 0 0
> 1339164300 0 0 0 0
>
> ***@lists.oetiker.ch
> https://lists.oetiker.ch/cgi-bin/listinfo/mrtg

[Daniel McDonald]

On 6/18/12 7:49 AM, "Attila Rajmund Nohl" <***@gmail.com> wrote:

> Hello!
>
> This is an absolute measure. It's memory usage of a process and is
> only updated every 15 minutes.

That's not the meaning of absolute, it is most likely a gauge.

Absolute resets the counter to zero after polling. It is displayed as a
rate after dividing by 300 (or whatever the polling interval is). There is
no counter rollover taken into account.

Counter assumes a continuously monotonically increasing value with
roll-overs. It is expressed as a rate.

Gauge is a snapshot of data that should be displayed as is and not
considered as a rate.

--
Daniel J McDonald, CCIE # 2495, CISSP # 78281


>
> 2012/6/9 Steve Shipway <***@auckland.ac.nz>:
>> Possibly you should be using the 'gauge' option?  Is this variable a counter,
>> or an absolute measure?
>>
>> Also the rate of change may be < 1 per sec.  If this is the case, native mode
>> MRTG cannot cope with it, you need RRDTool as your backend.
>>
>> Steve
>>
>> Steve Shipway
>> University of Auckland ITS
>> UNIX Systems Design Lead
>> ***@auckland.ac.nz
>> Ph: [phone] ext 86487
>>
>>
>> ________________________________________
>>
>> Now I don't get errors, but the data is not updated either. This is
>> how the first few lines of the <target>.log file look like:
>>
>> 1339164901 340201472 340201472
>> 1339164901 0 0 0 0
>> 1339164611 0 0 0 0
>> 1339164600 0 0 0 0
>> 1339164300 0 0 0 0
>>
>> ***@lists.oetiker.ch
>> https://lists.oetiker.ch/cgi-bin/listinfo/mrtg
>
> _______________________________________________
> mrtg mailing list
> ***@lists.oetiker.ch
> https://lists.oetiker.ch/cgi-bin/listinfo/mrtg

[Attila Rajmund Nohl]

2012/6/18 Daniel McDonald <***@austinenergy.com>:
>
>
>
> On 6/18/12 7:49 AM, "Attila Rajmund Nohl" <***@gmail.com> wrote:
>
>> Hello!
>>
>> This is an absolute measure. It's memory usage of a process and is
>> only updated every 15 minutes.
>
> That's not the meaning of absolute, it is most likely a gauge.
>
> Absolute resets the counter to zero after polling.  It is displayed as a
> rate after dividing by 300 (or whatever the polling interval is).  There is
> no counter rollover taken into account.
>
> Counter assumes a continuously monotonically increasing value with
> roll-overs.  It is expressed as a rate.
>
> Gauge is a snapshot of data that should be displayed as is and not
> considered as a rate.

Now I've tried gauge, but my data is still not updated:

Target[eselivm3v80l_memory]:
1.3.6.1.4.1.193.105.1.1.1.21.1.0&1.3.6.1.4.1.193.105.1.1.1.21.1.0:***@eselivm3v80l:4000::::2
Title[eselivm3v80l_memory]: Memory usage of NETSim
MaxBytes[eselivm3v80l_memory]: 10000000
RouterUptime[eselivm3v80l_memory]: ***@eselivm3v80l:4000
Options[eselivm3v80l_memory]: growright,nopercent,unknaszero,gauge,noo,pngdate
LegendI[eselivm3v80l_memory]: Memory usage of NETSim, as reported by the OS
LegendO[eselivm3v80l_memory]: Memory usage of NETSim, as reported by the OS

And the .log file:
1340026600 338206720 338206720
1340026600 0 0 0 0
1340026500 0 0 0 0
1340026200 0 0 0 0
1340025900 0 0 0 0
1340025600 0 0 0 0
1340025300 0 0 0 0
1340025000 0 0 0 0
1340024700 0 0 0 0

[Daniel McDonald]

On 6/18/12 8:38 AM, "Attila Rajmund Nohl" <***@gmail.com> wrote:

> 2012/6/18 Daniel McDonald <***@austinenergy.com>:
>>
>> On 6/18/12 7:49 AM, "Attila Rajmund Nohl" <***@gmail.com> wrote:
>> Gauge is a snapshot of data that should be displayed as is and not
>> considered as a rate.
>
> Now I've tried gauge, but my data is still not updated:
>
> Target[eselivm3v80l_memory]:
> 1.3.6.1.4.1.193.105.1.1.1.21.1.0&1.3.6.1.4.1.193.105.1.1.1.21.1.0:***@eseli
> vm3v80l:4000::::2
> Title[eselivm3v80l_memory]: Memory usage of NETSim
> MaxBytes[eselivm3v80l_memory]: 10000000

Max bytes of 10 million
> RouterUptime[eselivm3v80l_memory]: ***@eselivm3v80l:4000
> Options[eselivm3v80l_memory]: growright,nopercent,unknaszero,gauge,noo,pngdate
> LegendI[eselivm3v80l_memory]: Memory usage of NETSim, as reported by the OS
> LegendO[eselivm3v80l_memory]: Memory usage of NETSim, as reported by the OS
>
> And the .log file:
> 1340026600 338206720 338206720
Latest value is 338 million, larger than maxbytes, so MRTG considers it
invalid....

Increase your maxbytes to a billion (or whatever the real amount of memory
available is) and it should work fine.


> 1340026600 0 0 0 0
> 1340026500 0 0 0 0
> 1340026200 0 0 0 0
> 1340025900 0 0 0 0
> 1340025600 0 0 0 0
> 1340025300 0 0 0 0
> 1340025000 0 0 0 0
> 1340024700 0 0 0 0


--
Daniel J McDonald, CCIE # 2495, CISSP # 78281

[Attila Rajmund Nohl]

2012/6/18 Daniel McDonald <***@austinenergy.com>:
[...]
> Latest value is 338 million, larger than maxbytes, so MRTG considers it
> invalid....
>
> Increase your maxbytes to a billion (or whatever the real amount of memory
> available is) and it should work fine.

Thanks, this was the problem, now I get good graphs.