Microsoft has made a number of Azure monitoring and management services generally available. You can now access Azure Monitor, Azure Network Watcher, Azure Resource Health, and Azure Advisor in general availability (GA).
It has been a busy period of time as a number of different management solutions became generally available in Microsoft Azure. Each of these solutions can be used separately but most, if not all, can be integrated into other services. An example of this is using Operations Management Suite (OMS) for consolidated management.
The goal of Azure Monitor is to give you essential monitoring capabilities without acquiring or configuring costly and timely third-party solutions. With Azure Monitor, you get platform-level and service-level telemetry.
Effective management starts with designing for management by exception. Failing IT managers want status displays to constantly be updating. Successful IT managers spend their time more wisely. They only want to know when things are not working as they should. You can get classic alerts based on thresholds. You can also get alerts based on activities such as virtual machine reboots, deployment failures, or permission changes.
Monitoring data sources include:
Monitoring data can be used as follows:
The network is critical for connecting components of a service but it is even more critical in cloud computing:
Network Watcher gives us the ability to monitor our network deployments in two ways but scenario-based monitoring provides us with:
You can use resource monitoring when a scenario identifies a trouble spot:
Microsoft describes Azure Advisor as a personalized cloud consultant. Advisor aims to help you with:
The recommendations are split into four categories. Each is assigned an impact level or severity,:
The key word in Microsoft’s description is consultant. As with all consultants, this tool will talk a ton and make lots of recommendations that are not necessarily correct for you. Advisor is useful because it is offering another pair of eyes on your deployment. You must be able to understand and filter the advice. You cannot treat each recommendation as an alert in Event Viewer. This would be a great way to increase costs.
As a cloud customer, you are isolated from the fabric and the infrastructure management tools that make your services possible. This can make it difficult to understand the root cause of misbehavior. You need to discern if there is software fault in the virtual machine or if the underlying host is at fault.
Everything that you deploy and is an object in a resource group (ARM) is a resource. Resource Health is decided based on a number of signals from the underlying infrastructure. Once a resource becomes unhealthy, Resource Health can analyze additional information to identify the root cause. With this knowledge, you know how to proceed:
You can find Resource Health in the management blade of every resource that is supported. The current and historical (up to 14 days) health of the resource can be seen. You can also view advice for solving a current problem.