If you were at the AIMS Bootcamp in Rotterdam on 30 October, you would have seen Ricardo Torre give a "sneak preview" of the new AIMS agent for Azure. Well, now those new features are available for everyone, and I'm excited to share the news with the integration community – AIMS has released a new agent that enables us to automagically monitor our Azure Services (and control Azure consumption costs) using artificial intelligence!
So what's so cool about the new AIMS Azure agent?
The new agent collects thousands of metrics and metadata from your Azure subscription and uses machine learning algorithms to gain insights into normal behavior. This way, it's able to predict future trouble with your Azure Services and you will be able to prevent downtime from happening. But, if you've used AIMS before, this shouldn't be a surprise, as these capabilities have been available for their other agents for years (BizTalk, SQL, SDK ++).
The Azure agent brings some new capabilities too – AIMS calls them Azure governance features – that help you stay in control of your Azure environment and costs:
First, the agent is able to automatically identify new resources and applications that are deployed in Azure. Us BizTalkers know that deploying a new resource or app on BizTalk could be quite time consumping, and Azure makes this fast and easy by comparison. This is great for productivity, but it means that companies can easily lose control of their Azure costs. By automatically detecting new deployments, AIMS helps you keep on top of things.
Second, AIMS helps you keep control of your Azure costs by tracking consumption on a much more granular level than what's available from Microsoft by default. When developing in Azure, a mistake is easily made. Most developers don’t think of keeping costs low, they just want to create beautiful functionality and get the job done. I have seen it happen a lot, developers playing in Azure and keeping LogicApps running, or companies forgetting to shut down the an Azure Virtual Machine over night or during weekends so it keeps running and consuming resources and money.
Azure gives us hourly consumption metrics out-of-the-box, and we can see our billing data on a daily basis only. If consumption spikes, it's easy to see how the costs can start increasing rapidly before we see the problem, giving us a nasty surprise. With AIMS, you can now see your Azure consumption metrics down to the minute, allowing you to prevent enormous bills!
Pretty cool huh?
Which Azure Services are supported?
AIMS is already covering a lot of Azure Services, and more will be added in time based on customer requests. Here is what's covered today:
- API Management
- Virtual Machines (Classic Compute, Compute/VMs, Compute/Disks)
- Network/Network Interfaces
- Network/Public IP Addresses
- Cosmos DB
- Event Hub
- Key Vault
- Logic Apps
- Service Bus
- Signal R Service
- Azure SQL DB
- Azure Storage
- App Service Plans
- Azure Web Apps
- Azure Functions, and;
- Azure billing data
How do I get started with the AIMS Azure agent?
You can get started with the Free Edition of AIMS from the AIMS website. After connecting the agent to your Azure Subscription (which typically takes under 10 minutes) the agent collects 1 week of historical data from Azure. This data allows AIMS machine learning algorithms to automatically build normal behavior patterns for every metric collected from each Azure Service.
Using these normal behavior patterns, AIMS identifies anomalies on each single metric (including consumption data) and correlates the anomalies across the rest of the metrics to generate anomaly warnings – which means you are up and running with predictive anomaly detection on Azure with hours.
Sounds easy, doesn't it. I can confirm that it is. Give it a try!
Topics from this blog: Blog