Previous

Next

Exercise - Use diagnostic data

Completed

Azure enables you to take action on the metrics you're capturing. You can create alerts that email you, or start automations with Runbooks.

Now, let's create an alert based on your web server's CPU usage. If the CPU usage goes above the set threshold of 90 percent for five minutes, you get an email.

In this unit, you add an alert for your virtual machine (VM). Next, you cause the VM CPU usage to spike, and receive the alert email.

Create an alert based on CPU usage

If CPU usage goes above 90 percent for one minute, you'd like to get a notification.

1. Go to the Azure portal.

2. On the Azure portal menu, or from the Home page, select Virtual machines.

3. Select the monitored-linux-vm virtual machine that you created. The monitored-linux-vm virtual machine pane appears.

4. On the left menu pane, scroll down to Monitoring, and select Alerts. The Alerts pane appears for the monitored-linux-vm virtual machine.

5. On the top menu, select Create > Alert rule.

   

The Create an alert rule pane appears with the Condition tab open.

Select the resource

When you open the Alerts pane from the menu of your virtual machine, the resource for your alert rule is automatically set to that virtual machine.

To verify, select the Scope tab on the Create an alert rule pane, and confirm that monitored-linux-vm is listed under Resource.

To set a different resource, you would choose Select scope and make the setting in the Select a resource pane.

Create the condition

1. On the Create an alert rule pane, open the Condition tab.

2. In the Signal name search box, search for and select Percentage CPU.

3. Under Alert logic, confirm or enter the following values.

   Setting	Value
   Threshold	Static
   Aggregation type	Average
   Operator	Greater than or equal to
   Threshold value	90

   

Add an action

1. On the Create an alert rule pane, open the Actions tab and select Create action group. The Create action group pane appears.

2. On the Basics tab, enter the following values for each setting.

   Setting	Value
   Project details
   Subscription	Your subscription
   Resource group	Resource group that contains the VM
   Region	Global
   Instance details
   Action group name	Email alerts
   Display name	Email alerts

   

3. Select Next: Notifications, and enter the following values for each setting.

   Setting	Value
   Notifications
   Notification type	Email/SMS/Push/Voice
   Name	High CPU usage

   The Email/SMS message/Push/Voice pane appears.

4. Tick the Email checkbox, and enter your email address to receive the alert notification. If this pane doesn't appear, select Edit for the notification type.

   

5. On the Email/SMS message/Push/Voice pane, select OK.

6. On the Create an action group pane, select Review + create, and then select Create. You should see a notification that your action group was successfully created.

Enter alert rule details

1. On the Create an alert rule pane, select the Details tab and enter the following values for each setting.

   Setting	Value
   Alert rule name	Web server CPU running hot
   Alert rule description	Email alert notifies when the CPU is over 90%

   

2. Select Review + create, and then select Create. You should see a notification that your Alert rule was successfully created.

Trigger the alert

1. In the left menu pane, go to the Overview pane for the monitored-linux-vm virtual machine.

2. In the upper menu bar, select Connect, and then select SSH. The Connect pane for your VM appears.

   

3. Under step 4 on the Connect page, copy the azureuser@ and IP address that follows.

4. Open Azure Cloud Shell (Bash), enter SSH, and paste what you copied in the previous step. It should look like the following command.

   ssh azureuser@<ipaddress>
   

   If Cloud Shell has timed out, select Reconnect.

5. When you're prompted, enter yes to connect to the VM.

6. Run the following command to update the list of available updates.

   sudo apt-get update
   

7. Run the following command to install the stress tool on the VM.

   sudo apt-get install stress
   

8. Run the following command to stress the VM's CPU.

   sudo stress --cpu 16 -v -t 10m
   

9. Return to the Azure portal and in the upper left corner, select the ☰ icon, and then select Dashboard. The KPI Dashboard appears.

10. On the Max cpu/usage_active for monitored-linux-vm graph, select the ellipses (...) on the upper right, and then select Refresh.

11. After a few minutes, the CPU should approach 100 percent, and you'll receive an alert email.

    

Continue