Trying to Monitor a Azure local VM using Azure Monitor but i am not able to see any metric even after enabling the insights

Question

Trying to Monitor a Azure local VM using Azure Monitor but i am not able to see any metric even after enabling the insights

VIshnu 80

Hi Team,

I am trying to Monitor a virtual machine that is deployed on a Azure local using azure monitor but i am not able to see the metrics

User's image

Due to which i am unable to see any metric signals in the alerts configuration.

User's image

VIshnu 80 Reputation points

2025-06-19T11:35:55.6166667+00:00

FYI,

ARC agent is connected and Monitoring agent[agent version :1.53.03069.2298] is also installed
Mounika Reddy Anumandla 6,495 Reputation points Microsoft External Staff Moderator

2025-06-19T12:31:48.5+00:00
Hi VIshnu,

In order for Azure Monitor to work on a machine, it needs to be associated with a Data Collection Rule (DCR). You include the resource ID of the DCR when you create your policy definition.
Azure Arc machines (unlike native Azure VMs) do not automatically emit platform metrics. They rely on:

AMA + DCR + Log Analytics + (optional) Metrics-from-logs setup
This article reviews recommended deployment methods for the Azure Monitor agent VM extension, across multiple production physical servers or virtual machines in your environment, to help you determine which works best for your organization.
https://learn.microsoft.com/en-us/azure/azure-arc/servers/azure-monitor-agent-deployment#deploy-the-extension-individually-on-each-machine
https://learn.microsoft.com/en-us/azure/azure-arc/servers/deploy-ama-policy

Step 1: Create a Log Analytics Workspace

If not already created:

Go to Azure Portal → Log Analytics workspaces → + Create

Region: Match the region of the Arc resource if possible

Resource Group: vishnurg (from your image)

Step 2: Create a DCR (Data Collection Rule)

You can use the Azure Portal:

Go to Azure Monitor → Data Collection Rules → + Create

Choose:

Platform: Windows

Add performance counters:

\Processor(_Total)\% Processor Time

\LogicalDisk(_Total)\% Free Space

\Memory\% Committed Bytes In Use

Destination:

Select your Log Analytics Workspace

Assign the Arc Machine (visnutest1) as a target for this rule.

Hope it helps!

Let me know if you have any further queries!
If the comment is helpful, please click "upvote" to let us know!
VIshnu 80 Reputation points

2025-06-20T05:17:11.12+00:00

Hello Mounika,
We have already enabled the DCR and Log analytics

I could see that performance counters are added by default

But when i try to re add the 3 counters mentioned above i get below error
VIshnu 80 Reputation points

2025-06-20T05:25:11.25+00:00

Also I see the performance insights as well, but not the able to see in the metric tab which is blocking me to setup alerts
Mounika Reddy Anumandla 6,495 Reputation points Microsoft External Staff Moderator

2025-06-20T09:35:24.66+00:00
VIshnu,

Thanks for the screenshots — they provide great clarity.

You're seeing Insights data under the "Performance" tab, which means the Azure Monitor Agent (AMA) and Data Collection Rule (DCR) are working for Insights.

However:

You're unable to see performance metrics under the "Metrics" blade to configure alerts.

When adding counters like \VmInsights\DetailedMetrics, you're hitting this error:

CounterSpecifiers' predefined counter categories '\VmInsights\DetailedMetrics' cannot be mixed with other specifiers.

Azure Monitor doesn't support mixing custom performance counters with \VmInsights\DetailedMetrics in the same DCR. These counters are predefined and reserved for VM Insights.

That’s why:

Metrics blade isn’t showing the expected metrics.

You cannot create alerts from the Metrics blade as no standard metric namespace is available (only Logs and Insights are working). Explanation:

The UI for data collection rules understands only certain data collections you can enable like performance counters, windows logs and syslogs. Thus, when you browse data collection rules for VM Insights it cannot show its configuration. Also, when you try to add things like performance counters to VM Insights data collection rule it tells you that you should not mix these two data collections together. The best practice is to have separate data collection rules for different sets of data.

For example, you will have one data collection rule for VM Insights, another one for Change tracking, third one for performance counters and windows events, fourth one for security events, etc. You can have multiple data collection rules and do a separate assignment for each rule to each VM. So, create new data collection rule for performance counters for example and assign it to the VMs you want these performance counters to be collected. Azure Monitor agent picks all configurations that are assigned and sends the data appropriately.

Metrics blade depends on metrics being sent to the Azure Monitor Metrics database. The DCR with \VmInsights\DetailedMetrics sends data only to Log Analytics, not to Metrics. You need to configure a separate DCR explicitly targeting Azure Monitor Metrics to see counters in Metrics blade.

Once your custom DCR is linked:

Go to Metrics blade of the Arc machine.

Select "Azure Monitor Metrics" namespace.

You'll see performance counters available.

You can now create metric alerts!

Hope it helps!

Please feel free to tag me for further queries!
VIshnu 80 Reputation points

2025-06-20T11:38:55.69+00:00

I have created a DCR for Platform Metrics and selected the destination as "Azure Monitor Metrics (preview)" where as my previous DCR is "Azure Monitor logs"

Below image is for Azure Local VM

so does that mean that for logs and Metrics i need to use different DCR's unlike in public azure VM that one DCR like below is pulling the metrics and Logs as well for my VM, but not working for VM deployed in Azure local(ARC enabled one)

Below Image is for a public azure VM

Ask
1.is there any public document where its mentioned that we need to have a separate DCR for the Metrics for azure local or arc enabled VM's

2.I could see the for Azure local VM the Performance encounter metrics in Preview , is there any feature to collect the metric which is in GA or only this is the option.

3.what are the recommended alerts the Microsoft recommends for a VM [As i do not see any recommended alerts in the alerts blade for azure local VM, which we see in public azure]
Mounika Reddy Anumandla 6,495 Reputation points Microsoft External Staff Moderator

2025-06-20T12:49:53.1033333+00:00

Hi VIshnu,

While there's no single document stating it outright in one sentence, Microsoft’s official documentation does clearly outline the behavior and constraints.
The documentation indicates that the two types of counters are assigned to different data destinations—Logs vs Metrics—supporting the need for two DCRs.
https://learn.microsoft.com/en-us/azure/azure-monitor/vm/data-collection-performance
For guidance on creating a DCR to collect performance counters, see Collect events and performance counters from virtual machines with Azure Monitor Agent. You can quickly create a DCR by using the most common counters. For more granular filtering by criteria such as event ID, you can create a custom filter by using XPath queries.

Collecting performance counters via Azure Monitor Agent (AMA) and sending them to Azure Monitor Metrics is still in preview for Azure Arc-enabled servers There's no GA alternative yet—the only way today is the AMA + DCR + Metrics preview combo. https://learn.microsoft.com/en-us/azure/azure-monitor/vm/data-collection-performance?
There are no auto-generated recommended alerts for Arc/local VMs—you need to create them manually using a mixture of metric and log alerts.
https://gbbcore.wordpress.com/2024/12/12/where-are-my-recommended-alerts-for-arc-enabled-servers

Hope it helps!

Let me know if you have any further queries!
If the comment is helpful, please click "upvote" to let us know!
Mounika Reddy Anumandla 6,495 Reputation points Microsoft External Staff Moderator

2025-06-23T01:39:58.26+00:00

Hi VIshnu,
Just checking to see if you had a chance to review the response provided to your question!

If you have any further queries, please let me know.

If the information is helpful, please "upvote" to let us know.
VIshnu 80 Reputation points

2025-06-23T04:24:34.7666667+00:00
Hello Mounika,

Thank you for the update. I am able to see the guest metrics [even though they are slightly different than Public azure]

So now i wanted to set below 4 alerts

If my CPU is more than 85%

If my memory is utilized more than 80%

If my disk free space is less than 15% [if my keep \LogicalDisk(_Total)% Free Space then does this check for all the attached disk for the VM or only OS disk ]

How do get the alert if my VM is down due to any reason like availability metrics
Mounika Reddy Anumandla 6,495 Reputation points Microsoft External Staff Moderator

2025-06-23T05:00:36.7233333+00:00
Hi VIshnu,

As per my understanding, Azure Arc connects external or local machines to Azure, so everything relies on agents (AMA) and what they can collect from within the OS only. Some metrics like Memory % Used or Disk Read IOPS with 1-minute granularity are enhanced by the Azure VM platform, which is absent in Arc/local. You are limited to what the guest OS exposes through performance counters, not hypervisor/host data.

Alert when CPU > 85%

Metric: \Processor(_Total)\% Processor Time

Condition: Greater than 85

Aggregation: Average

Frequency: 5 minutes (typical)

Action: Email or action group as needed

_Total covers all CPU cores — this gives you the total average utilization across the system.

https://learn.microsoft.com/en-us/archive/technet-wiki/12984.understanding-processor-processor-time-and-process-processor-time

Alert when Memory utilization > 80%

Since we don’t get % used memory directly, use this approach:

Metric: \Memory\Available MBytes

Logic: Trigger alert when Available MBytes < (Total RAM * 20%).

Alert when Disk Free Space < 15%

Metric: \LogicalDisk(_Total)\% Free Space

Condition: Less than 15

Aggregation: Average or Minimum

✅ _Total means the combined view of all mounted disks (OS + data). If you want to target specific drives (like C: or D:), choose them explicitly (e.g., \LogicalDisk(C:)\% Free Space).

https://learn.microsoft.com/en-us/answers/questions/1294956/azure-metrics-alert-threshold-value-question
https://stackoverflow.com/questions/73405750/azure-monitor-alert-for-cpu-usage-80-for-30-minutes-or-more

Hope it helps!

Let me know if you have any further queries!

If the comment is helpful, please click "upvote" to let us know!
VIshnu 80 Reputation points

2025-06-23T05:14:37.41+00:00

can you help me with below queries
1.In the metrics for i don't see anything like \LogicalDisk(C:)% Free Space, i could only see total.

2.How do get the alert if my VM is down due to any reason like availability metrics

Accepted answer

1 additional answer

Your answer

VIshnu 80 Reputation points

2025-06-19T11:35:55.6166667+00:00

FYI,

ARC agent is connected and Monitoring agent[agent version :1.53.03069.2298] is also installed
Mounika Reddy Anumandla 6,495 Reputation points Microsoft External Staff Moderator

2025-06-19T12:31:48.5+00:00

Hi VIshnu,

In order for Azure Monitor to work on a machine, it needs to be associated with a Data Collection Rule (DCR). You include the resource ID of the DCR when you create your policy definition.
Azure Arc machines (unlike native Azure VMs) do not automatically emit platform metrics. They rely on:

AMA + DCR + Log Analytics + (optional) Metrics-from-logs setup
This article reviews recommended deployment methods for the Azure Monitor agent VM extension, across multiple production physical servers or virtual machines in your environment, to help you determine which works best for your organization.
https://learn.microsoft.com/en-us/azure/azure-arc/servers/azure-monitor-agent-deployment#deploy-the-extension-individually-on-each-machine
https://learn.microsoft.com/en-us/azure/azure-arc/servers/deploy-ama-policy

Step 1: Create a Log Analytics Workspace

If not already created:

Go to Azure Portal → Log Analytics workspaces → + Create

Region: Match the region of the Arc resource if possible

Resource Group: vishnurg (from your image)

Step 2: Create a DCR (Data Collection Rule)

You can use the Azure Portal:

Go to Azure Monitor → Data Collection Rules → + Create

Choose:

Platform: Windows

Add performance counters:

\Processor(_Total)\% Processor Time

\LogicalDisk(_Total)\% Free Space

\Memory\% Committed Bytes In Use

Destination:

Select your Log Analytics Workspace

Assign the Arc Machine (visnutest1) as a target for this rule.

Hope it helps!

Let me know if you have any further queries!
If the comment is helpful, please click "upvote" to let us know!
VIshnu 80 Reputation points

2025-06-20T05:17:11.12+00:00

Hello Mounika,
We have already enabled the DCR and Log analytics

I could see that performance counters are added by default

But when i try to re add the 3 counters mentioned above i get below error
VIshnu 80 Reputation points

2025-06-20T05:25:11.25+00:00

Also I see the performance insights as well, but not the able to see in the metric tab which is blocking me to setup alerts
VIshnu 80 Reputation points

2025-06-20T11:38:55.69+00:00

I have created a DCR for Platform Metrics and selected the destination as "Azure Monitor Metrics (preview)" where as my previous DCR is "Azure Monitor logs"

Below image is for Azure Local VM

so does that mean that for logs and Metrics i need to use different DCR's unlike in public azure VM that one DCR like below is pulling the metrics and Logs as well for my VM, but not working for VM deployed in Azure local(ARC enabled one)

Below Image is for a public azure VM

Ask
1.is there any public document where its mentioned that we need to have a separate DCR for the Metrics for azure local or arc enabled VM's

2.I could see the for Azure local VM the Performance encounter metrics in Preview , is there any feature to collect the metric which is in GA or only this is the option.

3.what are the recommended alerts the Microsoft recommends for a VM [As i do not see any recommended alerts in the alerts blade for azure local VM, which we see in public azure]
Mounika Reddy Anumandla 6,495 Reputation points Microsoft External Staff Moderator

2025-06-20T12:49:53.1033333+00:00

Hi VIshnu,

While there's no single document stating it outright in one sentence, Microsoft’s official documentation does clearly outline the behavior and constraints.
The documentation indicates that the two types of counters are assigned to different data destinations—Logs vs Metrics—supporting the need for two DCRs.
https://learn.microsoft.com/en-us/azure/azure-monitor/vm/data-collection-performance
For guidance on creating a DCR to collect performance counters, see Collect events and performance counters from virtual machines with Azure Monitor Agent. You can quickly create a DCR by using the most common counters. For more granular filtering by criteria such as event ID, you can create a custom filter by using XPath queries.

Collecting performance counters via Azure Monitor Agent (AMA) and sending them to Azure Monitor Metrics is still in preview for Azure Arc-enabled servers There's no GA alternative yet—the only way today is the AMA + DCR + Metrics preview combo. https://learn.microsoft.com/en-us/azure/azure-monitor/vm/data-collection-performance?
There are no auto-generated recommended alerts for Arc/local VMs—you need to create them manually using a mixture of metric and log alerts.
https://gbbcore.wordpress.com/2024/12/12/where-are-my-recommended-alerts-for-arc-enabled-servers

Hope it helps!

Let me know if you have any further queries!
If the comment is helpful, please click "upvote" to let us know!
Mounika Reddy Anumandla 6,495 Reputation points Microsoft External Staff Moderator

2025-06-23T01:39:58.26+00:00

Hi VIshnu,
Just checking to see if you had a chance to review the response provided to your question!

If you have any further queries, please let me know.

If the information is helpful, please "upvote" to let us know.
VIshnu 80 Reputation points

2025-06-23T04:24:34.7666667+00:00

Hello Mounika,

Thank you for the update. I am able to see the guest metrics [even though they are slightly different than Public azure]

So now i wanted to set below 4 alerts

If my CPU is more than 85%

If my memory is utilized more than 80%

If my disk free space is less than 15% [if my keep \LogicalDisk(_Total)% Free Space then does this check for all the attached disk for the VM or only OS disk ]

How do get the alert if my VM is down due to any reason like availability metrics
Mounika Reddy Anumandla 6,495 Reputation points Microsoft External Staff Moderator

2025-06-23T05:00:36.7233333+00:00

Hi VIshnu,

As per my understanding, Azure Arc connects external or local machines to Azure, so everything relies on agents (AMA) and what they can collect from within the OS only. Some metrics like Memory % Used or Disk Read IOPS with 1-minute granularity are enhanced by the Azure VM platform, which is absent in Arc/local. You are limited to what the guest OS exposes through performance counters, not hypervisor/host data.

Alert when CPU > 85%

Metric: \Processor(_Total)\% Processor Time

Condition: Greater than 85

Aggregation: Average

Frequency: 5 minutes (typical)

Action: Email or action group as needed

_Total covers all CPU cores — this gives you the total average utilization across the system.

https://learn.microsoft.com/en-us/archive/technet-wiki/12984.understanding-processor-processor-time-and-process-processor-time

Alert when Memory utilization > 80%

Since we don’t get % used memory directly, use this approach:

Metric: \Memory\Available MBytes

Logic: Trigger alert when Available MBytes < (Total RAM * 20%).

Alert when Disk Free Space < 15%

Metric: \LogicalDisk(_Total)\% Free Space

Condition: Less than 15

Aggregation: Average or Minimum

✅ _Total means the combined view of all mounted disks (OS + data). If you want to target specific drives (like C: or D:), choose them explicitly (e.g., \LogicalDisk(C:)\% Free Space).

https://learn.microsoft.com/en-us/answers/questions/1294956/azure-metrics-alert-threshold-value-question
https://stackoverflow.com/questions/73405750/azure-monitor-alert-for-cpu-usage-80-for-30-minutes-or-more

Hope it helps!

Let me know if you have any further queries!

If the comment is helpful, please click "upvote" to let us know!
VIshnu 80 Reputation points

2025-06-23T05:14:37.41+00:00

can you help me with below queries
1.In the metrics for i don't see anything like \LogicalDisk(C:)% Free Space, i could only see total.

2.How do get the alert if my VM is down due to any reason like availability metrics

Answer 1

Hi VIshnu,

Sorry if I misunderstood your question! The performance counter \LogicalDisk(_Total)\% Free Space provides an aggregate value across all logical disks (OS disk + data disks), not individual drives.

If my disk free space is less than 15% [if my keep \LogicalDisk(_Total)% Free Space then does this check for all the attached disk for the VM or only OS disk ]

This checks all the Total - All disks combined.

One of the most common monitoring requirements for a virtual machine is to create an alert if it stops running. The best method is to create a metric alert rule in Azure Monitor by using the VM availability metric. For a walk-through on this metric, see Create availability alert rule for Azure virtual machine.
Metric alert rules

A metric called Heartbeat is included in each Log Analytics workspace. Each virtual machine connected to that workspace sends a heartbeat metric value each minute. Because the computer is a dimension on the metric, you can fire an alert when any computer fails to send a heartbeat. Set the Aggregation type to Count and the Threshold value to match the Evaluation granularity.

Log search alert rules

Log search alerts use the Heartbeat table, which should have a heartbeat record every minute from each machine.

Use a rule with the following query:

Heartbeat | summarize TimeGenerated=max(TimeGenerated) by Computer, _ResourceId | extend Duration = datetime_diff('minute',now(),TimeGenerated) | summarize MinutesSinceLastHeartbeat = min(Duration) by Computer, bin(TimeGenerated,5m), _ResourceId

https://learn.microsoft.com/en-us/azure/azure-monitor/vm/monitor-virtual-machine-alerts#machine-availability

Hope it helps!

Answer 2

VIshnu 80

Hello Mounika,

Thank you for the support

Mounika Reddy Anumandla 6,495 Reputation points Microsoft External Staff Moderator

2025-06-24T04:55:24.5866667+00:00

Hi VIshnu,

Please take a moment to Accept answer and upvote it as a token of appreciation!

Accepted answer will help other community members navigate to the appropriate solutions.

Share via

Trying to Monitor a Azure local VM using Azure Monitor but i am not able to see any metric even after enabling the insights

1 additional answer

Your answer