Microsoft has a long history of turning its internal tools into products, especially within Azure. Since the company runs the same cloud-native apps as its customers, it faces similar challenges — just a bit ahead of the curve. A big part of managing those apps is keeping an eye on virtual machines (VMs). Whether they’re part of your infrastructure or powering platform features, VMs are everywhere in Azure. The scale of Azure lets apps grow fast and span the globe, but it makes tracking everything tricky. You might not know exactly how many VMs are running at any moment, but you want to ensure your apps stay stable and users can access services smoothly. Managing all that telemetry data from different VMs in various locations can be overwhelming.

Introducing Project Flash for Better VM Oversight

Microsoft has been working on a set of tools called Project Flash to help manage Azure VMs more easily. These tools aim to bring all your data into a single dashboard and categorize events to automate responses. Although still in development, some features are mature enough to try out now, with a mix of full releases and previews. The tools include enhancements to the Azure Portal, like the Resource Graph and Resource Health, plus previews of VM metrics in Azure Monitor. They also offer labeled events you can use with Azure Event Grid, making it easier to react to changes.

Making Sense of VM Data with Azure Resource Graph

One of the key parts of Project Flash is the Azure Resource Graph. It helps you investigate VM failures and downtime using a familiar query language called Kusto Query Language. You can track VM availability over the past two weeks, which helps with debugging and understanding what changes might have affected performance. The data is stored in the Resource Graph’s HealthResources table, allowing you to build custom dashboards in the Azure Portal. Once you find the right query, you can even integrate it into your own DevOps tools via the REST API. This way, you can keep a close eye on your VMs’ health and spot issues before they turn into bigger problems.

Using Event Grid to Automate VM Management

As your applications grow, managing VM health info across many instances becomes even more important. Traditional monitoring tools might struggle to keep up when VMs spin up and down dynamically, like in Virtual Machine Scale Sets. Microsoft built a publish-and-subscribe messaging system into Azure, which is perfect for this. It lets you send health notifications from remote systems to a central location, filter those messages, and trigger automated actions. Using Event Grid with Project Flash, you can route notifications about VM status changes to endpoints like Event Hubs or Azure Monitor. For example, if a VM goes offline, you can set up an alert to notify your team or even automatically spin up a replacement.

New Features and Future Developments for Better Insights

Microsoft is continuously adding new features to Project Flash. Recently, it announced enhancements for VM metrics in Azure Monitor, especially around VM availability. Now, metrics include labels like platform, customer, or unknown, which helps pinpoint the cause of issues quickly. This makes troubleshooting faster because you can see if a problem stems from Azure’s infrastructure or your application. Another exciting update is the integration of Event Grid with Azure Monitor. It allows event data from multiple sources to flow into your core monitoring system, enabling real-time alerts via SMS or push notifications. Additionally, Microsoft plans to provide more detailed insights into the physical infrastructure, such as rack-level hardware issues or predicted failures. This can give you advance warning to migrate workloads or perform maintenance.

In the end, Project Flash aims to make managing virtual machines more transparent and automated. Its evolving features help us understand how cloud automation impacts our apps. By combining these tools with others like Azure’s monitoring services, users can build custom dashboards and automation workflows. One innovative idea is merging Project Flash’s Cloud Events with tools like Drasi, which could lead to a more advanced control plane that oversees complex platforms and applications. This would give platform operators better control, quicker responses, and more reliable services, even as cloud environments grow more complicated.

Inspired by

• https://www.infoworld.com/article/4031677/managing-azure-vms-with-project-flash.html

Sources

• gmpg.org
• networkworld.com
• microsoft.com
• microsoft.com
• microsoft.com