Share via

Storage metrics not populating for PostgreSQL Flexible Server - known platform issue

Askardeen Buhari 20 Reputation points
2026-01-22T01:30:07.22+00:00

Issue Summary

Storage metrics (storage_percent, storage_used, storage_free) have not been collecting data for my PostgreSQL Flexible Server since creation 1+ months ago, despite having identical configuration to a working dev server.

Problem Description

All three storage-related metrics return zero data points: - storage_percent - storage_used - storage_free

Verification Commands


*# Check for data points (returns 0)* 

az monitor metrics list \ --resource "/subscriptions/.../***-db" \ --metric "storage_percent" \ --start-time "2025-11-17T00:00:00Z" \ --query "value[0].timeseries[0].data[?average != null] | length(@)" Result: 0 
Azure Database for PostgreSQL

Answer accepted by question author

Pilladi Padma Sai Manisha 10,110 Reputation points Microsoft External Staff Moderator
2026-01-23T10:30:15.5066667+00:00

Hi Askardeen Buhari,
The issue occurred due to a code defect in the Azure PostgreSQL Flexible Server monitoring pipeline that introduced a race condition in the component responsible for publishing customer-visible storage metrics. Because of this condition, the storage-related metrics (storage_percent, storage_used, and storage_free) were never successfully emitted for the affected server, even though the server itself was healthy and correctly configured. This is why metric queries consistently returned zero data points from the time of server creation, while an identically configured dev server was unaffected.

Mitigation and Resolution

Resolving this issue required manual backend intervention to repair the metric emission state on the affected PostgreSQL Flexible Server. Once the correction was applied, the monitoring pipeline was able to process and publish storage metrics normally. No configuration changes were required from the customer side.

Current Status

After the backend fix, storage metrics are expected to populate going forward. Historical data prior to the correction cannot be backfilled, but new metric data should now appear correctly in Azure Monitor.

Was this answer helpful?

0 comments No comments

2 additional answers

Sort by: Most helpful
  1. Askardeen Buhari 20 Reputation points
    2026-01-22T01:31:51.38+00:00

    How to create azure support ticket through azure portal?

    Was this answer helpful?


  2. AI answer

    2026-01-22T01:30:16.5066667+00:00

    It appears that you are experiencing an issue with the storage metrics (storage_percent, storage_used, storage_free) not populating for your PostgreSQL Flexible Server. This can happen for several reasons, including configuration issues or potential platform-related bugs.

    Here are some steps you can take to troubleshoot the issue:

    1. Verify Metric Definitions: You have already checked if the metrics are defined, which returned "YES". This indicates that the metrics should be available for your server.
    2. Check Resource Configuration: Ensure that the configuration of your PostgreSQL Flexible Server matches the working dev server. Sometimes, subtle differences in configuration can lead to discrepancies in metric collection.
    3. Azure Support: If the metrics are defined correctly and the configurations are identical, it may be worth reaching out to Azure support. They can provide insights into whether this is a known issue or if there are any ongoing platform problems affecting metric collection.
    4. Monitoring Tools: Consider using Azure Monitor or other monitoring tools to see if they provide any additional insights or data points that might not be visible through the command line.
    5. Wait for Data Points: Sometimes, metrics can take time to start populating, especially if the server was recently created. If the server is still within a short timeframe since creation, it may be worth waiting a bit longer to see if data starts to appear.

    If the issue persists, documenting your findings and any error messages when contacting support will be beneficial.

    AI-generated content may be incorrect. Read our transparency notes for more information.

    Was this answer helpful?

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.