Metrics

Metrics are resource values collected at particular points in time, indicating system performance for monitored services or events. You define the interval at which these numerical values are captured, along with associated values that include the data and timestamp, the name of the metric, and labels that help define the metric. Metrics help reveal trends that you can analyze to identify performance bottlenecks or areas that require deeper analysis for a resource.

Azure uses a time-series database to store and optimize metrics, as this type of database is best suited to host this kind of data. Once combined with other data types, such as logs, you can better identify root causes of issues. Metrics are lightweight data points and therefore are ideal for use in Dashboards for real-time health monitoring of resources that are critical for your environment. (See Figure 4-2.)

Log data

Every system, including VMs, storage, network infrastructure, and other Azure services, generates logs that capture changes, issues, and errors. Logs can contain numerical or text data. In the context of Azure, you can store logs in a Log Analytics workspace (see Figure 4-3) and use Azure Monitor to analyze them for specific events or errors that are of interest. You can also perform queries against log data to identify the root cause of issues or to pre-empt issues for remedial action. Finally, you can set up alerts against logged data that help you proactively respond to issues before users are affected.

Distributed traces

A distributed trace is when a system traces a user request across a distributed environment and is used to identify and analyze any bottlenecks or other issues in your environment. Distributed systems contain a number of individual components; distributed traces help correlate logs generated by these individual components in the application stack. You use the Application Insights SDK to enable distributed tracing and define where to store trace data. You can then analyze trace data, along with other log data, and create dashboards, custom queries, and alerts for monitoring.

Curated visualization

Curated visualizations are limited in scope to a particular service or set of services. You can use them without having to configure them first. Examples of curated visualizations include the metrics that become visible when you click a specific Azure resource such as a VM (see Figure 4-4) or Azure Storage account (see Figure 4-5).

Insights

Insights are larger scalable visualizations that focus on specific resources, such as web apps, VMs, and container services. These require integration and customization but provide deeper and richer insights. Microsoft provides several Insights tools, including Application Insights, Container Insights, VM Insights, Network Insights, Storage Insights, and many others. Four Insights that can help you monitor your environment include the following:

• Application InsightsApplication Insights provides a deep understanding of an integrated web application’s performance, availability, usage, bottlenecks, and exceptions. The web application can be hosted anywhere, including on-premises and other cloud environments. You use the Application Insights SDK to enable Application Insights configuration for your web applications and configure it to monitor and stream the required parameters to Azure Monitor for analysis. Application Insights integrates with Visual Studio to support DevOps scenarios.

• VM InsightsVM Insights enables you to monitor the performance and health of numerous VMs at once and to identify dependencies on external endpoints that impact the performance of a particular VM. You can use VM Insights for VMs hosted in Azure, on-premises, or in other cloud environments.

• Network InsightsNetwork Insights (see Figure 4-6) monitors the health of the various network endpoints running in your environment, including VPN gateways, VPN connections, private endpoints, virtual networks, and so on. You can also use Network Insights to monitor connection health and traffic flow logs for the different network security groups (NSGs).

• Storage InsightsStorage Insights monitors the health of the various storage accounts running in your Azure subscriptions. With Storage Insights, you can easily monitor key metrics, such as the following:

  • Transactions

  • Transactions Timeline

  • E2E Latency

  • Server Latency

  • Client Errors

You can also monitor the capacity of various storage accounts to proactively identify any that are reaching their maximum thresholds. Figure 4-7 shows a Storage Accounts Insights dashboard.

Dashboards

Azure enables you to build customized dashboards using different data sources and data types and display them in the Azure Portal for easy viewing. You can also share your custom dashboards with other Azure users, providing a consistent view of critical resources across your IT team. Essentially, when creating a dashboard, you select tiles that contain the different data types from the different data sources for display. You can then size them and align them based on your requirements.

Figure 4-8 shows a sample dashboard that contains the following information:

• All the resources present in the tenant

• Average CPU credits consumed by some VMs

• Average memory available in bytes for those VMs

• Average CPU percentage usage for those VMs

• User sign-in graph for the past few days

• Secure Score metric

• Service Health and Marketplace shortcut tiles

Workbooks

You can use Azure Workbooks to generate rich and interactive visual reports from multiple data sources. You can base these reports on publicly available templates that have been pre-configured to contain key metrics or start with a blank canvas and create your own. You can customize, save, and reuse templates as necessary. Figure 4-9 shows a list of the templates that are publicly available at the time of this writing, while Figure 4-10 shows the output of the Performance Analysis template.

Power BI

Microsoft has integrated Power BI with Azure Monitor. So, you can easily import logs from Azure Monitor into Power BI to create rich, interactive visualizations. A full-fledged business analytics service, Power BI offers extensive capabilities to create queries, custom dashboards, and reports based on your organizational needs.

Third-party integrations

Azure Monitor integrates with several types of third-party solutions. For example:

• IT service-management (ITSM) solutionsYou can connect Azure to your ITSM tool to automatically create tickets and to manage the lifecycle of the ticket. The Microsoft site contains an extensive list of supported ITSM products, which continues to grow as more integrations come online.

• Azure Monitor partner integrationsAzure Monitor supports integration with partner solutions to either extend capabilities provided by Azure Monitor or export logs stored in Azure Monitor to their systems.

Using The Azure Portal

To set up and use Azure Monitor using the Azure Portal, follow these steps.

Create Metrics Alerts

1. Log in to the Azure Portal, and browse to the VM you created in your Azure environment for the purposes of this walkthrough.

2. In the left pane of the VM’s configuration blade, under Monitoring,click Metrics. (See Figure 4-11.)

   

   A default metrics visualization opens in the right pane. (See Figure 4-12.) This visualization contains the following options:

   • Add MetricClick this button to add metrics to the visualization.

   • Line ChartClick this button and choose a different view for the visualization from the drop-down list that appears. The options are Area Chart, Bar Chart, Scatter Chart, and Grid.

   • Drill Into LogsClick this button to dig deeper into the logs captured for this resource for a better understanding of the metrics that you are viewing. (This feature is currently in preview, so we will not dive into it here, as its functionality is subject to change.)

   • New Alert RuleClick this button to set up a new alert rule for this metric. You will learn how to do this in a moment.

   • Save to DashboardClick this button to pin this metric visualization to a dashboard or send it to a workbook.

   • MetricsClick this to change the metric displayed in the visualization. The metric shown in Figure 4-12 is Source VM CPU Credits Consumed Avg.

     

3. Click the New Alert Rule button.

   The Create an Alert Rule wizard starts with the Condition tab displayed. (See Figure 4-13.) Depending on your needs, you can do one of the following:

   • Leave the settings in the Conditiontab as is.

   • Click the Add Conditionbutton and follow the prompts to add a new condition for the rule to assess along with the current one.

   • Click the link for the condition in the Condition Namelist to edit the condition.

For this exercise, we will edit the condition.

4. Click the link for the condition in the Condition Name list.

   The Alert Logic section of the Configure Signal Logic dialog box offers different options depending on what you select as the Threshold setting. We will explore these different settings now.

5. In the Alert Logic section, under Threshold, click Static.

   You’ll see the settings shown in Figure 4-14. You don’t need to change any of these settings; we are simply going to review them here:

   • OperatorThe options in this drop-down list are Greater Than, Less Than, Greater Than or Equal To, and Less Than or Equal To.

   • Aggregation TypeThe options in this drop-down list are Average, Maximum, Minimum, Total, and Count.

   • Threshold ValueHere, you enter the threshold value to monitor against.

   • UnitThis is where you select the threshold unit.

   • Aggregation Granularity (Period)This defines the period over which the aggregation points will be grouped for analysis.

   • Frequency of EvaluationThis is where you specify how often the rule should be triggered.

   

6. In the Alert Logic section, under Threshold, click Dynamic.

   You’ll see the settings shown in Figure 4-15. For this exercise, we will create a dynamic threshold, so you will change these settings as outlined here.

   • OperatorThe options in this drop-down list are the same as for the Static Threshold settings. For this example, choose Greater or Less Than.

   • Aggregation TypeThe options in this drop-down list are the same as for the Static Threshold settings. For this example, choose Average.

   • Threshold SensitivityThe options in this drop-down list are High, Medium, and Low. For this example, choose Medium.

   • Aggregation Granularity (Period)This option is the same as for the Static Threshold settings. For this example, choose 5 Minutes.

   • Frequency of EvaluationThis option is the same as for the Static Threshold settings. For this example, choose Every 5 Minutes.

   

7. Click Advanced Settings, enter the following information, and click OK. (See Figure 4-16.)

   • Number of ViolationsThis number specifies how many violations should occur before the alert is triggered. For this exercise, choose 4.

   • Evaluation PeriodThis indicates the timeframe during which the specified number of violations must occur to trigger an alert. In this case, choose 20 Minutes.

   • Ignore Data BeforeTo begin monitoring for the alert conditions at some date and time in the future, select this check box and choose the desired date and time. We’ll leave this unchecked in this case.

   

8. Back in the Condition tab of the Create an Alert Rule wizard (see Figure 4-17), click Next.

   

   In the Actions tab of the Create an Alert Rule wizard (see Figure 4-18), you can select an existing action group or create a new one. In our case, we will create a new one.

   

9. Click the Create Action Group button.

10. In the Basics tab of the Create an Action Group wizard (see Figure 4-19), enter the following information. Then click Next.

    • SubscriptionSelect the subscription you want to use to create the action group.

    • Resource GroupSelect the resource group you want to use to create the action group. Alternatively, click the Create New link and follow the prompts.

    • Action Group NameType a unique name for the action group.

    • Display NameType a unique display name for the action group. This can contain a maximum of 12 characters.

    

    The Notifications tab opens. (See Figure 4-20.) Here, you select the notification type for any alerts triggered by the role. You can choose to send the notification by email to an Azure resource manager role or to send it by email, SMS message, push, or voice to the recipient(s) of your choosing.

    

11. In the Notification Type drop-down list, choose Email/SMS Message/Push/Voice.

12. In the Email/SMS Message/Push/Voice dialog box (see Figure 4-21), enter the following information. Then click OK.

    • EmailSelect this check box and enter the email address to send email notifications to when the alert rule is triggered.

• SMSSelect this check box and enter a country code and phone number to send SMS notifications to when the alert is triggered.

• Azure Mobile App NotificationIn this case, leave this unchecked. You would select this if you wanted to send a notification by way of an Azure mobile app.

• VoiceIn this case, leave this unchecked. You would select this if you wanted to send a voice-based notification.

• Enable the Common Alert SchemaYou select this option to standardize the alert schema across all the different alerting options. In this case, click No.

13. In the Notifications tab of the Create an Action Group wizard, click Next.

14. In the Actions tab, open the Action Type drop-down list and choose what type of action you want to perform when the alert rule is triggered. In this case, choose Automation Runbook. Then enter a name for the action type (here, RestartVM) and click Next. (See Figure 4-22.)

    

    

15. In the Configure Runbook dialog box (see Figure 4-23), enter the following information. Then click OK.

    • Run RunbookClick Enable to specify that the runbook should be run.

    • Runbook SourceSelect the runbook source. In this case, choose Built-in.

    When you choose Built-in as the runbook source, the dialog box displays additional settings. These are also shown in Figure 4-23.

    • RunbookChoose the built-in runbook you want to use. In our case, we select Restart VM.

    • SubscriptionChoose the Azure subscription that contains the Azure Automation account you want to use.

    • Automation AccountSelect the automation account you want to use. Alternatively, click New and follow the prompts.

    • Enable the Common Alert SchemaFor this example, choose No.

    

16. In the Actions tab of the Create an Action Group wizard, click Next.

17. In the Tags tab (see Figure 4-24), enter a name and value for any tags you want to associate with the action group, and click Next.

    

18. In the Review + Create tab (see Figure 4-25), review your settings, and click Review + Create to create the action group.

    

    The action group you just created appears in the Actions tab of the Create a Metric Alert wizard (see Figure 4-26).

    

19. Optionally, create additional action groups.

20. Click Next.

21. In the Details tab of the Create a Metric Alert wizard (see Figure 4-27), enter the following information. Then click Next:

    • SubscriptionSelect the subscription you want to use to create the metric alert.

    • Resource GroupSelect the resource group you want to use to create the metric alert. Alternatively, click the Create New link and follow the prompts.

    • SeverityChoose a severity for the alert when the rule conditions are met. You can select between Critical, Error, Warning, Informational, and Verbose based on your requirements. Here, select 1-Error.

    • Alert Rule NameType a unique name for the alert rule.

    • Alert Rule DescriptionType a description of the alert rule.

    • Enable Upon CreationSelect this check box.

    • Automatically Resolve AlertsSelect this check box.

    

22. In the Tags tab (see Figure 4-28), enter a name and value for any tags you want to define for the alert rule. Then click Next.

    

23. In the Review + Create tab (see Figure 4-29), review your settings and click Review + Create to create the metric alert.

    

Create Log Analytics Workspace

Next, you need to create a Log Analytics workspace.

24. Type log analytics workspace in the Azure Portal search box and select it from the list that appears. (See Figure 4-30.)

    

25. Click the Create Log Analytics Workspace button. (See Figure 4-31.)

    

26. In the Basic tab of the Create Log Analytics Workspace wizard (see Figure 4-32), enter the following information. Then click Next:

    • SubscriptionSelect the subscription you want to use to create the workspace.

    • Resource GroupSelect the resource group you want to use to create the workspace. Alternatively, click the Create New link and follow the prompts.

    • NameEnter a unique name for the workspace.

    • RegionSelect the Azure region to host the workspace.

    

27. In the Tags tab (see Figure 4-33), type a name and value for any tags you would like to associate with the workspace, and click Next.

    

28. In the Review + Create tab (see Figure 4-34), review your settings, and click Review + Create to create the workspace.

    

Configure Azure VM Monitoring

Now that you have created the Log Analytics workspace, you can set up monitoring for your Azure VM.

29. In Azure Portal, navigate to the VM for which you want to set up monitoring.

30. In the left pane of the VM’s configuration blade, under Monitoring, click Metrics. (See Figure 4-35.)

    

31. In the right pane, click the Enable button to enable Insights. (See Figure 4-36.)

    

32. Azure Monitor automatically locates an existing Log Analytics workspace and its associated subscription. If it is not the right one for your environment, select the appropriate subscription and Log Analytics workspace. (See Figure 4-37.)

    

33. Click the Enable button.

    When the VM’s integration with the Log Analytics workspace is complete, the VM’s Performance tab opens with the Logical Disk Performance view displayed. This view includes the following metrics for VM resources by default (see Figure 4-38):

    • CPU Utilization %

    • Available Memory %

    • Logical Disk IOPS

    • Logical Disk MB/s

34. Click the Map tab.

You will see a map similar to the one shown in Figure 4-39. This map contains all the active ports on the VM and the processes actively using each port.

35. Click a port or service to see more information about it. (See Figure 4-40.) This will give you deeper insights into the VM.

    

Set Up Azure VM Insights Alerts

At this point, you’re ready to set up an alert. For this example, you will create an alert for low disk space. Because your test VM likely does not have low disk space at this time, you can either set the monitoring threshold very high or add data to the VM disk to reduce the space below your monitoring threshold. In our case, we will do the former—raise an alert if the C drive goes below 90% free. (In production use, you would generally not set such a high threshold for disk space alerts.)

36. In the pane to the right of the Map tab, click the Log Events link.

37. In the Event Type list (see Figure 4-41), click InsightsMetrics.

    

38. The Log Explorer window opens with a pre-defined query that searches for all InsightsMetrics for the VM. (See Figure 4-42.) You will see various metrics, including Network, LogicalDisk, Processor, Memory, and others.

39. Locate the metric for which you want to set up the alert—in our case, the LogicalDisk FreeSpacePercentage metric.

    

40. In the pre-populated query at the top of the Log Explorer window, enter the following code after the VM name (‘SourceVM’), as shown in Figure 4-43. Then click Run.

    Click here to view code image

    And Namespace == 'LogicalDisk' and Name == 'FreeSpacePercentage'

    

41. The Log Explorer window returns a list of LogicalDisk FreeSpacePercentage metrics. (See Figure 4-44.)

    

42. Click the New Alert Rule button along the top of the Log Explorer window. (See Figure 4-45.)

    

43. In the Condition tab of the Create an Alert Rule wizard, under Measurement, change the following settings (see Figure 4-46):

    • MeasureChoose Val.

    • Aggregation TypeSelect Average.

    • Aggregation GranularityChoose 5 Minutes.

    

44. Leave the settings in the Split by Dimensions section of the Condition tab at their default values.

45. In the Alert Logic section of the Condition tab (see Figure 4-47), enter the following information. Then click Next:

    • OperatorSelect Less Than or Equal To.

    • Threshold ValueType 90. (Remember: You want a high value here to make it easier to trigger the alert, just for the purposes of this exercise.)

    • Frequency of EvaluationSelect 5 Minutes.

    

46. In the Actions tab, click Add Action Groups, and select the action group you created earlier. Alternatively, you can click Create Action Group and follow the prompts to create a new one. (See Figure 4-48.) Then click Next.

    

47. In the Details tab (see Figure 4-49), change the following settings (leave the others at their default values). Then click Next:

    • SubscriptionSelect the subscription you want to use to create the alert rule.

    • Resource GroupSelect the resource group you want to use to create the alert rule. Alternatively, click the Create New link and follow the prompts.

    • SeveritySelect 2-Warning.

    • Alert Rule NameType a unique name for the alert rule.

    • Alert Rule DescriptionType a description of the alert rule.

    • RegionSelect the Azure region in which to create the alert rule.

    • Enable Upon CreationSelect this check box.

      

48. In the Tags tab (see Figure 4-50), type a name and value for any tags you want to add for this alert rule, and click Next.

    

49. In the Review + Create tab, review your settings, and click Review + Create to create the alert rule.

50. After the alert has been created, let your system run for 15 or 20 minutes.

    The alert should trigger frequently, indicating low disk space on the test VM. (See Figure 4-51.)

    

Configure Azure Storage Monitoring

Finally, you can set up monitoring for an Azure Storage account.

51. In Azure Portal, navigate to the Azure Storage account for which you would like to set up monitoring.

52. In the left pane of the Azure Storage account’s configuration blade, under Monitoring, click Diagnostic Settings. (See Figure 4-52.)

    

    The pane on the right shows the current diagnostic status of the various storage types. (See Figure 4-53.)

    

53. Click the blob entry in the Name list.

    Azure displays the blob’s diagnostic configuration. As shown in Figure 4-54, the blob’s diagnostic settings are currently not configured.

    

54. Click the Add Diagnostic Setting link.

55. In the Diagnostic Setting dialog box (see Figure 4-55), change the following settings. (Leave any other settings at their default value.) Then click Save.

    • Diagnostic Setting NameType a name for the diagnostic setting.

    • Logs/CategoriesSelect the check box next to each log category you want to collect—in this case, StorageRead, StorageWrite, and StorageDelete.

    • MetricsSelect the check box next to each metric you want to collect—here, Transaction.

    • Send Log to Analytics WorkspaceSelect this check box. Then enter the details for the workspace in the settings that appear, including the workspace’s subscription and name. In this case, select the workspace you created earlier.

    

    You can set up the diagnostic settings for any other storage resource type by following these same steps. (See Figure 4-56.) You can apply the same settings as you did for blob or different settings based on the storage type. You can also set up diagnostic settings on the storage account level to apply the same settings across the entire account.

Run Queries On Azure Storage

Now that you’ve configured the diagnostic settings for log collection for your storage account, run a few queries in the Azure Portal to monitor the output and set up alerts if needed.

56. In the left pane of the configuration blade for the Azure Storage account for which you configured the diagnostic settings, under Monitoring, click Logs. (See Figure 4-57.)

    

    A query explorer opens, with a set of pre-defined queries for you to run. These include frequently used queries for auditing, performance monitoring, error tracking, and alerting purposes. (See Figure 4-58.)

    

    Because you most likely would not have any errors or performance bottlenecks in your test storage, you need to select a query that might show some output, such as one that generates a Frequent Operations chart.

57. Click the Run button in the Frequent Operations tile of the query explorer.

    The system displays a pie chart showing the frequency of all operations performed on this test storage account. (See Figure 4-59.)