Atlassian Team members are employees working across the company in a wide variety of roles.
January 9, 2020 edited
Hey folks, I'm a former webMethods (SoftwareAG) developer and I am currently Head of the Statuspage and Opsgenie Customer Success Team here at Atlassian. I wanted to follow up and share some steps that came out of a Proof-Of-Concept activity that my team conducted recently.
In the past, I received questions from SoftwareAG developers that seemed to follow a pattern:
There was an important problem occurring within a developed solution which required someone to be notified of the issue as soon as possible. Preferably, while the issue was still occurring, but definitely as near real-time as possible.
The problems are sometimes complex type failure conditions where a simple System LogError type activity was insufficient to address the problem.
The ability to capture and include contextual information in the notification would be fantastic.
In a few cases, actual FTE’s were being assigned to monitor for, and take action, whenever the problem occurred.
A simple SMS, or email, notification was helpful. However, actually waking someone up and providing actionable data about the problem - as near real-time as possible - was preferred.
Being able to track and understand the frequency of occurrence/recurrence over time would be helpful.
With those in mind, I came up with a general recommendation: Go ahead and connect your Digital Transformation Solution (webMethods in this particular case) to Opsgenie and have the tool create Alerts within Opsgenie. Add in all the extra information you’d need into the Alert and leverage the Notification, Escalation, Logging and Reporting capabilities of Opsgenie. I hope the video below provides the rest of the context to allow you to replicate the configuration.
Please note, the information here still only outlines how to connect the products and get the messages flowing between them. If someone eventually wants to create a fully flexible and reusable package/module/etc, in the respective tools, then by all means - please, go for it! For now, I’m only looking to close that small instructional gap and lower the bar on what might have been perceived as a technical hurdle.
The high-level pieces to be mindful of boil down to:
How to connect and Authenticate from the Tool to Opsgenie.
How/Where to perform an HTTP GET/POST from within the Tool to Opsgenie.
How/Where to pass the “payload” (JSON) between the Tool and Opsgenie
I will not be addressing things like:
How to build the JSON payloads required, using the Tool.
How to resolve any issues (like pagination) which may be required in JSON payloads returned from Opsgenie.
