I'm currently working on a monitoring project that aims to gather messages from various sources in our customers' cloud environments, like Advisor and Monitor, and analyze them to proactively identify potential issues. Everything is implemented through a Python script. I'm exploring options for a platform to centralize this information and I've come across Azure DevOps. While I'm aware it's not typically considered the best solution for monitoring tasks, my company already uses it for other units. I want a platform that allows me to view messages, filter them by customers, and apply tags. Would Azure DevOps be a suitable choice for my needs?
4 Answers
You can technically use Azure DevOps to run your Python script, but it’s really more of a CI/CD repository than a dedicated monitoring solution.
Using Azure DevOps for monitoring? Now that's an interesting angle to explore!
I don't think Azure DevOps is the best fit, but it depends on what you're looking to do. It's more like GitHub. You could create tasks in it using the CLI or API, but ideally, this should come from whatever system you already use for managing user stories or incidents.
Azure DevOps can be useful for your project in a couple of ways, but calling it a complete solution might be a stretch. Your script can't run in Azure DevOps, but it could serve as a Git repository and CI/CD pipeline to deploy your script, say to a small Azure VM or a web app. You can then use the API to log your events as work items in Azure DevOps. It may seem a bit redundant, but if centralization is your goal, it could work well in keeping everything in one place. Plus, once the work items are in there, you can filter, query, and even create dashboards.
That's actually what I was thinking too! I might use an Azure function to run the script. I know there are better alternatives out there, but this setup seems to fit my requirements for now. Appreciate the insights!
I get it! So you're suggesting that, based on my script's analysis, I could generate tasks in Azure DevOps? I'm just trying to wrap my head around it, especially with the API calls you mentioned.