Anyone who operates systems wants to know whether they are running, whether they are well configured, and whether a problem is on their side or Azure's. Azure provides three tools for this, which you get to know in this last lesson: Azure Advisor gives recommendations, Azure Service Health shows the state of Azure itself, and Azure Monitor collects and analyzes data.
The example is the made-up online pharmacy Lindenhof, which runs its shop in Azure and does not want to be woken up at night unless it is necessary.
Azure Advisor: recommendations for your resources
Azure Advisor evaluates your Azure resources and makes recommendations to improve reliability, security, performance and cost efficiency. You can think of it as a personal advisor for best practices in the portal. Each recommendation contains an action that you can implement right away, postpone or dismiss. On request, Advisor notifies you when new recommendations appear.
The recommendations fall into five categories:
| Category | What it is about |
|---|---|
| Reliability | Keeping applications running, detecting configuration risks |
| Security | Detecting threats and vulnerabilities that could lead to breaches |
| Performance | Finding changes that speed up applications |
| Operational Excellence | Improvements for processes and deployment |
| Cost | Ways to reduce spending |
The dashboard shows the recommendations of all subscriptions, and you can filter by subscription, resource group or service.
Azure Service Health: is Azure doing badly, or are we?
Azure Service Health keeps you informed about the state of Azure and your resources. It combines three views that narrow from the global to the individual:
- Azure Status shows the state of all services and regions worldwide. You check here when you hear of a major outage and want to know whether it affects Azure.
- Service Health focuses on the services and regions you actually use. It shows outages, planned maintenance and notices that apply to your environment. You can set up alerts so that you are notified automatically.
- Resource Health looks at individual resources, for example a specific virtual machine. It shows whether the resource is running normally and whether a problem is on Azure's side or on yours.
Azure Monitor: collect data, view it, react
Azure Monitor is the platform for collecting data from your Azure resources and applications, analyzing it and reacting to it. It works for Azure, for on-premises environments and for multiple clouds. Azure Monitor collects logs and metrics from applications, operating systems and network layers, stores them centrally and makes them available through dashboards, queries and alerts. There are three parts you should know:
- Log Analytics: the tool in the portal where you write and run queries on the collected data. Simple: find all errors of the last hour. Demanding: analyze and display developments over time.
- Azure Monitor Alerts: they notify you when a condition you defined occurs. You create an alert rule with the condition and an action group that determines who is notified and what happens afterward. Alerts can be based on metrics or on logs. A metric alert, for example, sends an email when the CPU of a VM stays above 80 percent.
- Application Insights: a part of Azure Monitor that monitors the performance and usage of your web applications, whether they run in Azure, on premises or in another cloud. It captures, for example, request rate, response times and failure rates, calls to dependencies, page load times and user counts. In addition there are availability tests, which send artificial requests to your application so that you notice when it does not respond, even with little traffic.
Together: Advisor tells you what you could do better. Service Health tells you whether Azure is the cause. Monitor shows what is happening in your systems right now, and warns you.
The example of the online pharmacy Lindenhof
Advisor recommends configuring the database more reliably and points to a rarely used, expensive resource. Application Insights reports that the order page is getting slower. An alert rule sends an email to the on-call team when the failure rate rises. And when the shop is unreachable at night, Resource Health shows whether the cause lies with the virtual machine, and Service Health shows whether Azure has reported a disruption in the region.
Exercise 1
Which tool do you use for these questions?
- Are there resources in my subscription that I could run more cheaply or more securely?
- Is the West Europe region affected by a disruption right now, and is there maintenance scheduled for next week that affects my services?
- I want an email when the CPU load of a VM is too high for a longer time.
- I want to search the logs for all errors of the last hour.
- I want to know how fast my web application responds and how many requests fail.
Solution 1
- Azure Advisor with the categories Cost and Security.
- Azure Service Health. It shows outages and planned maintenance that affect the services and regions you use. Azure Status gives the worldwide overview.
- Azure Monitor Alerts: a metric alert on the CPU, plus an action group that sends the email.
- Log Analytics in Azure Monitor: a query on the collected logs.
- Application Insights: response times and failure rates of web applications.
What comes next
With that you have an overview of the topics of the skills that Microsoft lists for Azure Fundamentals. The depth for administrators and developers, with exercises of its own, is offered by the academy in the course K3, module 3.1 "Azure fundamentals."
Microsoft Learn also leads further. The three learning paths "Microsoft Azure Fundamentals: Describe cloud concepts," "Describe Azure architecture and services" and "Describe Azure management and governance" are in this course's list. There you find Microsoft's own detailed texts.