I think I'm missing a step creating the graph from a script

[allxrt]

I think I'm nearly there, I have a cronjob that every 5 mins feeds in some stats from a raspberry pi in to to a file, I have set up a short bash script that basically reads the last line of the file. The script takes an argument in this case it is the MAC address of the raspberry pi so reads the correct log file.

I think I have created most of the steps I need but can't see how to create the graph itself as there isn't a host as such any help appreciated.

File:extract_probe_stats.sh $1 is the filename (mac address) that I would like to pass in the <rpi_mac> field on the data input which I beleive I declare in the Templates > Data Source > Data Template > Custom Data Field

#!/bin/bash
tail -1 ~/pythonScripts/rpimanager/cactiDs/ds/$1.ds

This will return the following data I will use for the data input:

freq:2.437 sig:79 uploadSen:2.04 uploadRec:1.65 dnloadSen:2.04 dnloadRec:1.65

I have created my Data Input method where it calls the shell script with my argument as <rpi_mac>:



My input field looks like this:


This is my data source template which I have created for each of the Mac addresses used for the custom data:


Here are the data templates:


And finally the graph template I have created, I will need to create one of these for each of the Raspberry Pis I am collecting stats on:

[allxrt]

I will definitely create a howto guide, if someone can help me through this last step.

Please let me know if I need to explain my question better, I appreciate this might be a bit time consuming to read

[TheWitness]

Check your poller item table for entries, if using spine, run it like this as the apache account. It could be permissions, or you may have spine compiled to not source the environment, in which case, you need to use the full command, for example /bin/tail.

Code: Select all

./spine -R -V 3 -S -f device_id -l device_id

Replace device_id with the id of the device in question.

[allxrt]

Thanks for your reply, I've checked permissions and I'm not currently using spine, but I think what is different to my old deployment is that there doesn't appear to be an 'Add Data Source' under the management section, is this hidden or moved somewhere else? I think the issue is that I'm not sure how to use the created template on the new version.

This is my older exisiting version with the add button



This then takes me to this screen where I can specify the datasource template to use if it's not attached to a device



This is my latest deployment using a later version where I'm unable to find the add button(I'm guessing there is a new way to create a data source?)

[allxrt]

Actually I think I have found it this is where the new data source is created.

Although this doesn't show in my data sources list Management > Data Source?

I will re-check permissions, assuming once a data source is created as below it should active the script somewhere - I will try to active debug to see if it gives any messages

[allxrt]

Sorry for the constant barrage of messages.. shouldn't the data source I created show in the Troubleshooting > Data Sources section?

[allxrt]

After having another look the old data template did have a data source within it shown below, so that brings me back to whether I'm missing an add button under the managment > data sources section ?

[TheWitness]

The way I do it is create a Template and then instantiate it at the Device level. Much easier than hand tooling a Graph and Data Source. Scales better too...