Correlate data in Azure Data Explorer and Azure Resource Graph with data in a Log Analytics workspace

You can correlate data in Azure Data Explorer and Azure Resource Graph with data in your Log Analytics workspace and Application Insights resources to enhance your analysis in Azure Monitor Logs. Microsoft Sentinel, which also stores data in Log Analytics workspaces, supports cross-service queries to Azure Data Explorer but not to Azure Resource Graph. This article explains how to run cross-service queries from any service that stores data in a Log Analytics workspace.

Run cross-service queries by using any client tools that support Kusto Query Language (KQL) queries, including the Log Analytics web UI, workbooks, PowerShell, and the REST API.

Permissions required

To run a cross-service query that correlates data in Azure Data Explorer or Azure Resource Graph with data in a Log Analytics workspace, you need:

  • Microsoft.OperationalInsights/workspaces/query/*/read permissions to the Log Analytics workspaces you query, as provided by the Log Analytics Reader built-in role, for example.
  • Reader permissions to the resources you query in Azure Resource Graph.
  • Viewer permissions to the tables you query in Azure Data Explorer.

Query data in Azure Data Explorer by using adx()

Enter the identifier for an Azure Data Explorer cluster in a query within the adx pattern, followed by the database name and table.

adx('https://help.kusto.windows.net/Samples').StormEvents

Combine Azure Data Explorer cluster tables with a Log Analytics workspace

Use the union command to combine cluster tables with a Log Analytics workspace.

For example:

union customEvents, adx('https://help.kusto.windows.net/Samples').StormEvents
| take 10
let CL1 = adx('https://help.kusto.windows.net/Samples').StormEvents;
union customEvents, CL1 | take 10

Tip

Shorthand format is allowed: ClusterName/InitialCatalog. For example, adx('help/Samples') is translated to adx('help.kusto.windows.net/Samples').

When you use the join operator instead of union, you're required to use a hint to combine the data in the Azure Data Explorer cluster with the Log Analytics workspace. Use Hint.remote={Direction of the Log Analytics Workspace}.

For example:

AzureDiagnostics
| join hint.remote=left adx("cluster=ClusterURI").AzureDiagnostics on (ColumnName)

Join data from an Azure Data Explorer cluster in one tenant with an Azure Monitor resource in another

Cross-tenant queries between the services aren't supported. You're signed in to a single tenant for running the query that spans both resources.

If the Azure Data Explorer resource is in Tenant A and the Log Analytics workspace is in Tenant B, use one of the following methods:

  • Use Azure Data Explorer to add roles for principals in different tenants. Add your user ID in Tenant B as an authorized user on the Azure Data Explorer cluster. Validate that the TrustedExternalTenant property on the Azure Data Explorer cluster contains Tenant B. Run the cross query fully in Tenant B.
  • Use Lighthouse to project the Azure Monitor resource into Tenant A.

Connect to Azure Data Explorer clusters from different tenants

Kusto Explorer automatically signs you in to the tenant to which the user account originally belongs. To access resources in other tenants with the same user account, you must explicitly specify TenantId in the connection string:

Data Source=https://ade.applicationinsights.io/subscriptions/SubscriptionId/resourcegroups/ResourceGroupName;Initial Catalog=NetDefaultDB;AAD Federated Security=True;Authority ID=TenantId

Query data in Azure Resource Graph by using arg() (Preview)

Enter the arg("") pattern, followed by the Azure Resource Graph table name.

For example:

arg("").<Azure-Resource-Graph-table-name>

Here are some sample Azure Log Analytics queries that use the new Azure Resource Graph cross-service query capabilities:

  • Filter a Log Analytics query based on the results of an Azure Resource Graph query:

    arg("").Resources 
    | where type == "microsoft.compute/virtualmachines" and properties.hardwareProfile.vmSize startswith "Standard_D"
    | join (
        Heartbeat
        | where TimeGenerated > ago(1d)
        | distinct Computer
        )
        on $left.name == $right.Computer
    
  • Create an alert rule that applies only to certain resources taken from an ARG query:

    • Exclude resources based on tags – for example, not to trigger alerts for VMs with a “Test” tag.

      arg("").Resources
      | where tags.environment=~'Test'
      | project name 
      
    • Retrieve performance data related to CPU utilization and filter to resources with the “prod” tag.

      InsightsMetrics
      | where Name == "UtilizationPercentage"
      | lookup (
          arg("").Resources 
          | where type == 'microsoft.compute/virtualmachines' 
          | project _ResourceId=tolower(id), tags
          )
          on _ResourceId
      | where tostring(tags.Env) == "Prod"
      

More use cases:

  • Use a tag to determine whether VMs should be running 24x7 or should be shut down at night.
  • Show alerts on any server that contains a certain number of cores.

Create an alert based on a cross-service query from your Log Analytics workspace

To create an alert rule based on a cross-service query from your Log Analytics workspace, follow the steps in Create or edit a log search alert rule, selecting your Log Analytics workspace, on the Scope tab.

Note

You can also run cross-service queries from Azure Data Explorer and Azure Resource Graph to a Log Analytics workspace, by selecting the relevant resource as the scope of your alert.

Limitations

General cross-service query limitations

  • Cross-service queries don’t support parameterized functions and functions whose definition includes other cross-workspace or cross-service expressions, including adx(), arg(), resource(), workspace(), and app().
  • Cross-service queries support only ".show" functions. This capability enables cross-cluster queries to reference an Azure Monitor, Azure Data Explorer, or Azure Resource Graph tabular function directly. The following commands are supported with the cross-service query:
    • .show functions
    • .show function {FunctionName}
    • .show database {DatabaseName} schema as json
  • Database names are case sensitive.
  • Identifying the Timestamp column in the cluster isn't supported. The Log Analytics Query API won't pass the time filter.
  • Cross-service queries support data retrieval only.
  • Private Link (private endpoints) and IP restrictions don't support cross-service queries.
  • mv-expand is limited to 2,000 records.
  • Azure Monitor Logs doesn't support the external_table() function, which lets you query external tables in Azure Data Explorer. To query an external table, define external_table(<external-table-name>) as a parameterless function in Azure Data Explorer. You can then call the function using the expression adx("").<function-name>.

Azure Resource Graph cross-service query limitations

  • Microsoft Sentinel doesn't support cross-service queries to Azure Resource Graph.
  • When you query Azure Resource Graph data from Azure Monitor:
    • The query returns the first 1,000 records only.
    • Azure Monitor doesn't return Azure Resource Graph query errors.
    • The Log Analytics query editor marks valid Azure Resource Graph queries as syntax errors.
    • Joins are not supported when not using Hint.
    • These operators aren't supported: smv-apply(), rand(), arg_max(), arg_min(), avg(), avg_if(), countif(), sumif(), percentile(), percentiles(), percentilew(), percentilesw(), stdev(), stdevif(), stdevp(), variance(), variancep(), varianceif().

Combine Azure Resource Graph tables with a Log Analytics workspace

Use the union command to combine cluster tables with a Log Analytics workspace.

For example:

union AzureActivity, arg("").Resources
| take 10
let CL1 = arg("").Resources ;
union AzureActivity, CL1 | take 10

When you use the join operator instead of union, you need to use a hint to combine the data in Azure Resource Graph with data in the Log Analytics workspace. Use Hint.remote={Direction of the Log Analytics Workspace}. For example:

Perf | where ObjectName == "Memory" and (CounterName == "Available MBytes Memory")
| extend _ResourceId = replace_string(replace_string(replace_string(_ResourceId, 'microsoft.compute', 'Microsoft.Compute'), 'virtualmachines','virtualMachines'),"resourcegroups","resourceGroups")
| join hint.remote=left (arg("").Resources | where type =~ 'Microsoft.Compute/virtualMachines' | project _ResourceId=id, tags) on _ResourceId | project-away _ResourceId1 | where tostring(tags.env) == "prod"

Next steps