Continuing on from Part 1 where we explored pulling data from Weather Underground.
In this post we’ll look at how to use a service called Pushover. In their own words ‘Pushover is a service to receive instant push notifications on your phone or tablet from a variety of sources.’ I’ve used it in the past with perl and PHP scripts for my home automation system. I’ve had it send alerts on everything from home theatre PC TV episode alerts to temperature warnings in my kegerator to bitcoin prices on Coinbase. I lead an exciting life, really. It’s tightly integrated into IFTTT so it’s a really easy alerting engine to use with that service.
To recall where we left it in Part 1, we were using a nested workflow to pull JSON formatted data from Weather Underground and output a few specific variables to the error log.
To make that a little more useful, in this new workflow, Weather Underground 002, we’re going to add in another nested workflow to send that data to the Pushover web service. The schema looks like this:
We modify the attributes of this parent workflow to include the API Key for Pushover, User Key for Pushover, and the name of the vCO REST Operation. Pushover has you register a key for your application AND a key for your user. Presumably so one could re-use someone else’s application. We will now use this message variable to move data from one workflow to the other – it will contain the body of what we send to pushover.
Inputs and outputs for the parent workflow remain unchanged because we are only changing what happens inside of it. I did do a bit of cleanup on the embedded workflow while troubleshooting to remove the default parameters but you really don’t have to. The POST method threw me off for a bit but I got it working this way.
Registering the Pushover REST host is easy enough in vCO:
Registering the REST Operation is a little more complicated as this is a POST method instead of a GET:
Back to our workflows, we need to be sure we match up our bindings. Key is the message data that is coming as the output from the previous workflow:
Validate for any bugs and you should be good. Run this new workflow and you should see the same basic message that we created in the previous step. Yup still cold.
So that’s all fun and stuff, but what’s the real implication of it? Well, this is an extremely simple monitoring solution that doesn’t require and text message or email gateway. Any monitoring alert or error you want to trap in the vCO environment, you can have piped right to your devices via a canned workflow you can drop in.