Metrics for resource performance in Azure - Commands & Configuration

Introduction

When you run apps or services in the cloud, you want to know how well they are working. Metrics help you see if your resources like virtual machines or databases are healthy and fast. This helps you fix problems before users notice.

When you want to check if your virtual machine is using too much CPU or memory.

When you need to see how many requests your web app is handling per second.

When you want to monitor the disk space or network traffic of your cloud resources.

When you want to set alerts to get notified if a resource is not performing well.

When you want to analyze trends over time to plan for scaling your resources.

Commands

This command fetches the CPU usage percentage metric for the virtual machine named 'myVM' every 1 minute. It helps you see how busy the CPU is.

Terminal

az monitor metrics list --resource /subscriptions/00000000-0000-0000-0000-000000000000/resourceGroups/myResourceGroup/providers/Microsoft.Compute/virtualMachines/myVM --metric Percentage CPU --interval PT1M

Expected Output

{

  "cost": 0,

  "timeseries": [

    {

      "data": [

        {

          "timeStamp": "2024-06-01T12:00:00Z",

          "average": 15.3

        },

        {

          "timeStamp": "2024-06-01T12:01:00Z",

          "average": 18.7

        }

      ]

    }

  ]

}

→

--resource - Specifies the exact resource to get metrics from

→

--metric - Chooses which performance metric to retrieve

→

--interval - Sets how often to collect metric data

This command lists all available metrics you can monitor for the virtual machine 'myVM'. It helps you know what performance data you can collect.

Terminal

az monitor metrics list-definitions --resource /subscriptions/00000000-0000-0000-0000-000000000000/resourceGroups/myResourceGroup/providers/Microsoft.Compute/virtualMachines/myVM

Expected Output

{

  "value": [

    {

      "name": {

        "value": "Percentage CPU",

        "localizedValue": "Percentage CPU"

      },

      "unit": "Percent"

    },

    {

      "name": {

        "value": "Network In",

        "localizedValue": "Network In"

      },

      "unit": "Bytes"

    }

  ]

}

→

--resource - Specifies the resource to check available metrics

This command creates an alert named 'HighCPUAlert' that triggers when the average CPU usage of 'myVM' goes above 80%. Alerts help you react quickly to performance issues.

Terminal

az monitor metrics alert create --name HighCPUAlert --resource-group myResourceGroup --scopes /subscriptions/00000000-0000-0000-0000-000000000000/resourceGroups/myResourceGroup/providers/Microsoft.Compute/virtualMachines/myVM --condition "avg Percentage CPU > 80" --description "Alert when CPU usage is over 80%"

Expected Output

{

  "id": "/subscriptions/00000000-0000-0000-0000-000000000000/resourceGroups/myResourceGroup/providers/microsoft.insights/metricAlerts/HighCPUAlert",

  "name": "HighCPUAlert",

  "type": "microsoft.insights/metricalerts",

  "location": "global",

  "properties": {

    "description": "Alert when CPU usage is over 80%",

    "enabled": true

  }

}

→

--name - Names the alert rule

→

--scopes - Defines which resource the alert monitors

→

--condition - Sets the condition that triggers the alert

This command gets the number of requests received by the web app 'myWebApp' every 5 minutes. It helps you understand how much traffic your app is getting.

Terminal

az monitor metrics list --resource /subscriptions/00000000-0000-0000-0000-000000000000/resourceGroups/myResourceGroup/providers/Microsoft.Web/sites/myWebApp --metric Requests --interval PT5M

Expected Output

{

  "timeseries": [

    {

      "data": [

        {

          "timeStamp": "2024-06-01T12:00:00Z",

          "total": 120

        },

        {

          "timeStamp": "2024-06-01T12:05:00Z",

          "total": 135

        }

      ]

    }

  ]

}

→

--resource - Specifies the web app resource

→

--metric - Chooses the metric to retrieve

→

--interval - Sets the data collection frequency

Key Concept

If you remember nothing else from this pattern, remember: metrics show how your cloud resources are performing so you can keep them healthy and fast.

Common Mistakes

Using incorrect resource ID format in the command

The command fails because Azure cannot find the resource without the full correct ID

Always use the full resource ID starting with /subscriptions/ and including resource group and provider

Requesting a metric that is not available for the resource

The command returns an error or empty data because the metric does not exist for that resource type

Use 'az monitor metrics list-definitions' to check available metrics before querying

Not specifying the interval flag or using an invalid interval format

The command may return data points that are too sparse or the command fails due to wrong interval syntax

Use ISO 8601 duration format like PT1M for 1 minute or PT5M for 5 minutes

Summary

Use 'az monitor metrics list' with the full resource ID to get performance data like CPU or requests.

Check available metrics for a resource with 'az monitor metrics list-definitions' before querying.

Create alerts with 'az monitor metrics alert create' to get notified when performance crosses thresholds.

Practice

1. What is the main purpose of metrics in Azure resource performance monitoring?

easy

A. To create new Azure resources automatically

B. To change resource configurations

C. To delete unused resources

D. To track how well resources are working over time

Solution

1. Step 1: Understand what metrics represent

   Metrics provide data about resource usage and performance over time.

2. Step 2: Identify the purpose of monitoring metrics

   Monitoring helps keep resources healthy and plan for future needs.

3. Final Answer:

   To track how well resources are working over time -> Option D

4. Quick Check:

   Metrics = Performance tracking [OK]

Hint: Metrics measure performance, not resource creation or deletion [OK]

Common Mistakes:

• Confusing metrics with resource management actions
• Thinking metrics change resource settings
• Assuming metrics create or delete resources

2. Which Azure CLI command correctly lists metrics for a resource?

easy

A. az metrics get --resource-group

B. az resource metrics show --name

C. az monitor metrics list --resource

D. az monitor resource metrics fetch --id

Solution

1. Step 1: Recall the correct Azure CLI syntax for metrics

   The official command to list metrics is az monitor metrics list with the resource specified.

2. Step 2: Verify the options

   Only az monitor metrics list --resource uses the correct command and parameter format.

3. Final Answer:

   az monitor metrics list --resource <resource-id> -> Option C

4. Quick Check:

   Correct CLI command = az monitor metrics list --resource [OK]

Hint: Use 'az monitor metrics list' to get metrics data [OK]

Common Mistakes:

• Using incorrect command verbs like 'get' or 'show'
• Mixing resource and resource-group parameters wrongly
• Adding unsupported flags or commands

3. Given this Azure CLI command output snippet for CPU percentage metric:

{"value": [{"timeseries": [{"data": [{"timeStamp": "2024-06-01T10:00:00Z", "average": 30.5}, {"timeStamp": "2024-06-01T10:05:00Z", "average": 45.2}]}]}]}

What is the average CPU usage at 10:05 AM?

medium

A. 45.2%

B. No data available

C. 75.7%

D. 30.5%

Solution

1. Step 1: Locate the timestamp 10:05 AM in the data

   Find the entry with "timeStamp": "2024-06-01T10:05:00Z".

2. Step 2: Read the average CPU value at that time

   The average value is 45.2% at 10:05 AM.

3. Final Answer:

   45.2% -> Option A

4. Quick Check:

   CPU at 10:05 = 45.2% [OK]

Hint: Match timestamp exactly to find correct metric value [OK]

Common Mistakes:

• Picking the wrong timestamp value
• Adding values instead of reading single entry
• Confusing average with total or max

4. You run az monitor metrics list --resource <id> but get an error: "Resource not found". What is the most likely cause?

medium

A. Metrics are not enabled for the resource

B. The resource ID is incorrect or does not exist

C. You forgot to login to Azure CLI

D. The command syntax is wrong

Solution

1. Step 1: Understand the error message

   "Resource not found" means Azure cannot locate the resource by the given ID.

2. Step 2: Identify common causes

   An incorrect or non-existent resource ID is the most common reason for this error.

3. Final Answer:

   The resource ID is incorrect or does not exist -> Option B

4. Quick Check:

   Resource not found = Wrong ID [OK]

Hint: Check resource ID spelling and existence first [OK]

Common Mistakes:

• Assuming metrics must be enabled to avoid this error
• Ignoring login status
• Blaming command syntax without checking resource ID

5. You want to monitor the average disk read bytes for a virtual machine over the last 24 hours using Azure CLI. Which command correctly achieves this?

hard

A. az monitor metrics list --resource --metric DiskReadBytes --interval PT24H --aggregation Average

B. az monitor metrics list --resource --metric DiskReadBytes --interval 24h --aggregation Total

C. az monitor metrics show --resource --metric DiskReadBytes --time-range 24h --aggregation Average

D. az metrics list --resource --metric DiskReadBytes --interval PT24H --aggregation Average

Solution

1. Step 1: Identify the correct command and parameters

   The command to list metrics is az monitor metrics list. The metric name is DiskReadBytes. The interval uses ISO 8601 duration format (PT24H for 24 hours). Aggregation for average is 'Average'.

2. Step 2: Compare options for correctness

   az monitor metrics list --resource --metric DiskReadBytes --interval PT24H --aggregation Average uses correct command, metric, interval format, and aggregation. Others have wrong commands, interval formats, or aggregation types.

3. Final Answer:

   az monitor metrics list --resource <vm-resource-id> --metric DiskReadBytes --interval PT24H --aggregation Average -> Option A

4. Quick Check:

   Correct CLI + metric + interval + aggregation = az monitor metrics list --resource --metric DiskReadBytes --interval PT24H --aggregation Average [OK]

Hint: Use ISO 8601 interval format and 'Average' aggregation [OK]

Common Mistakes:

• Using wrong command like 'az metrics show'
• Using incorrect interval format like '24h'
• Choosing 'Total' instead of 'Average' aggregation