Multiple Data Sources.... Help!

dgarstang · Post by **dgarstang** » Sat Jun 07, 2008 5:42 pm

New cacti user.

I have a python script that generates the following output:

[root@mon02 voice]# ./vox_trunk_fadif.py --trunk-id=35
vox_trunk_attempted:6 vox_trunk_completed:7

For some reason cacti is not graphing the second value. I went and looked at the rrd file, and I see this:

<rrd> <version> 0003 </version>
<step> 300 </step> 
<lastupdate> 1212877811 </lastupdate> 

<ds>
<name> vox_trunk_completed </name>
<type> GAUGE </type>
<minimal_heartbeat> 600 </minimal_heartbeat>
<min> 0.0000000000e+00 </min>
<max> 1.0000000000e+02 </max>


<last_ds> U </last_ds>
<value> NaN </value>
<unknown_sec> 11 </unknown_sec>
</ds>

<ds>
<name> vox_trunk_attempted </name>
<type> GAUGE </type>
<minimal_heartbeat> 600 </minimal_heartbeat>
<min> 0.0000000000e+00 </min>
<max> 1.0000000000e+02 </max>


<last_ds> 6 </last_ds>
<value> 6.6000000000e+01 </value>
<unknown_sec> 0 </unknown_sec>
</ds>

 <rra>

As you can see, cacti is picking up a value of 'U' for the vox_trunk_completed data source. Why would it do that? Where is it getting 'U' from? I have a data template with the two data sources in it.

I can't figure out what's going on. I've deleted everything from the graph right down to the data input method and recreated multiple times and it keeps doing it.

Help!

dus001 · Post by **dus001** » Sun Jun 08, 2008 3:17 am

Maybe you should:
- check the associated data input methods (are both output fields names corrects) ?
- try update log level to show the result of the script execution in cacti.log
- add logs messages (into a file) directly into the called script ?

Post by **gandalf** » Sun Jun 08, 2008 3:19 am

Please make sure, both parameters are spit out by only one "print" statement. Cacti won't wait for the second one to be printed
Reinhard

dgarstang · Post by **dgarstang** » Sun Jun 08, 2008 11:03 am

Thanks.

I know my script is ok, because I went and changed the python script I had to just this:

#!/bin/sh

echo "vox_trunk_attempted:6 vox_trunk_completed:7"

Can't go wrong with that. I also went and compared everything (data input method, data template, graph template, data source, graph) side by side, screen by screen to another graph where, for some reason, I actually got it to work. As far as I can tell, they are the same.

I looked at the cacti logs. They aren't indicating any errors. I rebuilt the poller cache too.

What else could it be?

dus001 · Post by **dus001** » Sun Jun 08, 2008 12:08 pm

I still think you should:
- try update log level to show the result of the script execution in cacti.log

dgarstang · Post by **dgarstang** » Sun Jun 08, 2008 12:50 pm

Ok, I turned up the cacti debug. I see this:

06/08/2008 05:40:08 PM - CMDPHP: Poller[0] Host[81] DS[2058] CMD: /var/www/cacti/scripts/voice/vox_trunk_fadif.py --trunk-id=35, output: vox_trunk_attempted:6 vox_trunk_completed:7

That looks ok to me. If I compare it against the older one that works, it's output is:

06/08/2008 05:40:10 PM - CMDPHP: Poller[0] Host[81] DS[1996] CMD: /var/www/cacti/scripts/voice/vox_idd_fadif.py --idd-code=92, output: vox_idd_attempted:0 vox_idd_completed:0

Know that the id of rrd file on the one that isn't working is 2058, I also found:
06/08/2008 05:45:13 PM - CMDPHP: Poller[0] DEBUG: SQL Exec: "insert into poller_output (local_data_id, rrd_name, time, output) values (2058, '', '2008-06-08 17:45:11', 'vox_trunk_attempted:6 vox_trunk_completed:7')"

For the one that is working, the rrd id is 1975:
06/08/2008 05:45:14 PM - CMDPHP: Poller[0] DEBUG: SQL Exec: "insert into poller_output (local_data_id, rrd_name, time, output) values (1975, '', '2008-06-08 17:45:11', 'vox_idd_attempted:8 vox_idd_completed:2')"

The output format of these two data sources looks to be the same.

I did notice this difference however.

For the working one:
06/08/2008 05:46:49 PM - POLLER: Poller[0] CACTI2RRD: /usr/bin/rrdtool update /var/www/cacti/rra/voice_data_server_vox_idd_completed_1975.rrd --template vox_idd_attempted:vox_idd_completed 1212947111:8:2

For the non working one:
06/08/2008 05:46:08 PM - POLLER: Poller[0] CACTI2RRD: /usr/bin/rrdtool update /var/www/cacti/rra/voice_data_server_vox_trunk_completed_2058.rrd --template vox_trunk_attempted 1212947111:6

As you can see, for the non working one, cacti has dropped the vox_trunk_attempted variable. Why would it do that when it's getting the correct output from the script?

Actually, that last comparison made no sense. Now I'm confused.

dus001 · Post by **dus001** » Sun Jun 08, 2008 1:04 pm

The "vox_trunk_completed" appears in the used data input method as output field ?

I guess you've checked this plenty times, but sometimes...

dgarstang · Post by **dgarstang** » Sun Jun 08, 2008 1:11 pm

Yes, trunkid is an input field, and vox_trunk_attempted and vox_trunk_completed are both output fields. That's how they appear in the data input method. Is that what the output shows?

dus001 · Post by **dus001** » Sun Jun 08, 2008 1:19 pm

The output seems to indicate that the "vox_trunk_completed" parameter is not known as to be stored.

Argh..

"Update RRD File" is checked for both parameters in input method ?

dgarstang · Post by **dgarstang** » Sun Jun 08, 2008 1:22 pm

There is no 'Update RRD File' check box on the input variables, if that's what you meant.

However, both vox_trunk_completed and vox_trunk_attempted output fields in the data input method have 'Update RRD File' checked. There is only one input to the script, trunkid.

The command line to the Script/Cmd in the data input method is:
<path_cacti>/scripts/voice/vox_trunk_fadif.py --trunk-id=<trunkid>

trunkid is an input variable, vox_trunk_completed and vox_trunk_attempted are output variables.

dus001 · Post by **dus001** » Sun Jun 08, 2008 1:38 pm

Were the graph / datasources created only after the whole template finalization ?

If the response is no: I would suggest to test what is the result of a full fresh graph instantiation (either for the same server or another).

dgarstang · Post by **dgarstang** » Sun Jun 08, 2008 2:03 pm

I created the data input method first, followed by the data template, then the graph template, then the data source and finally the graph. I've repeated the whole process from scratch maybe half a dozen times and it's failed to work each time, except for that one before... which apparently I now can't repeat.

dgarstang · Post by **dgarstang** » Mon Jun 09, 2008 2:56 am

5 days later, and I still can't work out what's wrong with this and why it isn't working.

I started from scratch AGAIN, and cranked up the output. Here's all the relevant log lines I could find:

06/09/2008 07:50:07 AM - CMDPHP: Poller[0] Host[81] DS[2060] CMD: /var/www/cacti/scripts/voice/vox_trunk_fadif.py --trunk-id=35, output: vox_trunk_attempted:6 vox_trunk_completed:7

06/09/2008 07:50:07 AM - CMDPHP: Poller[0] DEBUG: SQL Exec: "insert into poller_output (local_data_id, rrd_name, time, output) values (2060, '', '2008-06-09 07:50:06', 'vox_trunk_attempted:6 vox_trunk_completed:7')"

06/09/2008 07:50:09 AM - CMDPHP: Poller[0] DEBUG: SQL Assoc: "select data_template_rrd.data_source_name, data_input_fields.data_name from (data_template_rrd,data_input_fields) where data_template_rrd.data_input_field_id=data_input_fields.id and data_
template_rrd.local_data_id=2060"

06/09/2008 07:50:10 AM - CMDPHP: Poller[0] DEBUG: SQL Exec: "delete from poller_output where local_data_id='2060' and rrd_name='' and time='2008-06-09 07:50:06'"

06/09/2008 07:50:11 AM - POLLER: Poller[0] CACTI2RRD: /usr/bin/rrdtool update /var/www/cacti/rra/voice_data_server_vox_trunk_attempted_2060.rrd --template vox_trunk_attempted 1212997806:6

..... Right there! Where the heck did vox_trunk_completed go???

Post by **gandalf** » Wed Jun 11, 2008 2:10 pm

Please post a screenshot of the failing data source associated with that host.
Reinhard

Multiple Data Sources.... Help!

Multiple Data Sources.... Help!

Who is online