In the DX APM Infrastructure agent, Azure Storage metrics—specifically "Used Capacity"—may intermittently disappear, show no values, or appear "grayed out" in the Investigator tree. This often occurs when monitoring both capacity-related metrics and transaction-related metrics simultaneously within the same Azure profile.
* DX APM Infrastructure Agent (APMIA)
* Azure Monitoring Extension
* Azure Storage Account integration
A granularity mismatch exists when both capacity-related metrics (typically requiring a 1-hour interval) and transaction-related metrics (often requiring a 5-minute interval) are defined within the same schema file. The agent struggles to reconcile these differing polling frequencies within a single schema profile, causing the data collection to fail or the metrics to become stale and eventually disappear.
To resolve this conflict, the Azure Storage metrics must be divided into two separate schema files based on their collection frequency. This configuration allows the agent to define independent polling intervals for each set of metrics.
Steps to Resolve:
Separate the Schema Files:
Ensure you are using two distinct schema files in your agent configuration:
azure_storage_schema.json: Used for capacity-related metrics (e.g., Used Capacity).azure_storage_transaction_schema.json: Used for transaction-related metrics.Update Agent Configuration:
Locate your agent configuration file (e.g., bundle.properties or your specific Azure profile configuration) and update the schemaList property to include both new files:
introscope.agent.azure.profiles.azure.1.schemaList=azure_cdn,azure_network_load_balancer,azure_storage,azure_storage_transaction,azure_vm
Configure Independent Polling Intervals:
Define the polling intervals separately for each schema to ensure efficient data collection.
introscope.agent.azure.profiles.azure.1.azure_storage.interval=3600
introscope.agent.azure.profiles.azure.1.azure_storage_transaction.interval=300
Restart the Agent:
Restart the DX APM Infrastructure Agent for the configuration changes to take effect.
After the restart, verify the Investigator tree. Both the capacity-related metrics and transaction metrics should now report data independently, and the "Used Capacity" metric should remain stable without graying out.
If metrics continue to be missing, check the APMIA debug logs for any errors related to these specific profile names.