Azure Database for PostgreSQL flexible server: Diagnostic log export to Log Analytics stopped suddenly for all categories, server healthy, recreating settings and restart have no effect

Hi @megamingus ,

Thank you for reaching out to Microsoft Q&A.

Based on the troubleshooting completed, this appears to be a resource-level diagnostic export issue, rather than a PostgreSQL engine, Log Analytics capacity, or destination-specific problem.

The main indicators are

All diagnostic categories stopped exporting within the same time window.

The PostgreSQL Flexible Server remains healthy and continues generating activity.

Deleting and recreating the diagnostic setting did not restore the export.

Restarting the server had no effect.

A separate diagnostic setting targeting an Azure Storage account also received no data.

The Log Analytics workspace shows no quota, throttling, or ingestion-related events.

Because both Log Analytics and Storage destinations are affected, the failure appears to occur before the logs reach either destination.

1. Can the customer inspect or reset the export pipeline

There is no documented customer-facing mechanism to inspect, restart, or reset the underlying diagnostic export pipeline for an individual PostgreSQL Flexible Server.

Recreating the diagnostic setting is the appropriate customer-side action. Since this has already been completed without restoring the export, Microsoft Support will need to investigate the backend state.

1. Should compute scaling be used as a workaround

A compute scale operation is not a documented or guaranteed remediation for diagnostic export failures. Scaling a production server solely to restore logging is not recommended, particularly before Support has reviewed the existing failure state.

1. Can Microsoft restore the export

Microsoft Support can investigate the resource-level diagnostic routing state and engage the appropriate Azure Database for PostgreSQL and Azure Monitor engineering teams if backend remediation is required.

The support request should include:

Subscription ID and server resource ID

Region: West Europe

Failure window: 24 August 2026, approximately 08:43–08:57 UTC

Diagnostic-setting configuration and resource ID

Log Analytics workspace and Storage account details

Last ingestion time for each affected table

Results of the diagnostic-setting recreation and server restart

Relevant Activity Log, Resource Health, and Service Health results

1. Can the missing logs be recovered

Logs that were not delivered to Log Analytics or Storage should not be assumed to be recoverable or replayable.

As an immediate step, check PostgreSQL Flexible Server → Monitoring → Server logs. PostgreSQL server log files may still be available for download if they remain within the configured retention period. However, this would not necessarily recover other categories such as Query Store Runtime, session statistics, autovacuum statistics, or PgBouncer logs.

For reference:

Configure and access logs in Azure Database for PostgreSQL Flexible Server

Logs in Azure Database for PostgreSQL Flexible Server

Diagnostic settings in Azure Monitor

In summary, the available evidence isolates the failure to the diagnostic export path for this resource. There is no documented customer-side reset, compute scaling is not a supported recovery method, and recovery of logs that were not exported cannot be guaranteed. The recommended next step is to open an Azure Support request for backend investigation and avoid further disruptive changes until the current state has been reviewed.

I hope this helps. Please let us know if you have any further questions.

If this answer helps, please consider clicking Accept Answer so that other community members facing similar issues can benefit from it as well.Thank you for reaching out to Microsoft Q&A.

Based on the troubleshooting completed, this appears to be a resource-level diagnostic export issue, rather than a PostgreSQL engine, Log Analytics capacity, or destination-specific problem.

The main indicators are:

All diagnostic categories stopped exporting within the same time window.

The PostgreSQL Flexible Server remains healthy and continues generating activity.

Deleting and recreating the diagnostic setting did not restore the export.

Restarting the server had no effect.

A separate diagnostic setting targeting an Azure Storage account also received no data.

The Log Analytics workspace shows no quota, throttling, or ingestion-related events.

Because both Log Analytics and Storage destinations are affected, the failure appears to occur before the logs reach either destination.

1. Can the customer inspect or reset the export pipeline

There is no documented customer-facing mechanism to inspect, restart, or reset the underlying diagnostic export pipeline for an individual PostgreSQL Flexible Server.

Recreating the diagnostic setting is the appropriate customer-side action. Since this has already been completed without restoring the export, Microsoft Support will need to investigate the backend state.

1. Should compute scaling be used as a workaround

A compute scale operation is not a documented or guaranteed remediation for diagnostic export failures. Scaling a production server solely to restore logging is not recommended, particularly before Support has reviewed the existing failure state.

1. Can Microsoft restore the export

Microsoft Support can investigate the resource-level diagnostic routing state and engage the appropriate Azure Database for PostgreSQL and Azure Monitor engineering teams if backend remediation is required.

The support request should include:

Subscription ID and server resource ID

Region: West Europe

Failure window: 24 August 2026, approximately 08:43–08:57 UTC

Diagnostic-setting configuration and resource ID

Log Analytics workspace and Storage account details

Last ingestion time for each affected table

Results of the diagnostic-setting recreation and server restart

Relevant Activity Log, Resource Health, and Service Health results

1. Can the missing logs be recovered

Logs that were not delivered to Log Analytics or Storage should not be assumed to be recoverable or replayable.

As an immediate step, check PostgreSQL Flexible Server → Monitoring → Server logs. PostgreSQL server log files may still be available for download if they remain within the configured retention period. However, this would not necessarily recover other categories such as Query Store Runtime, session statistics, autovacuum statistics, or PgBouncer logs.

For reference:

Configure and access logs in Azure Database for PostgreSQL Flexible Server

Logs in Azure Database for PostgreSQL Flexible Server

Diagnostic settings in Azure Monitor

In summary, the available evidence isolates the failure to the diagnostic export path for this resource. There is no documented customer-side reset, compute scaling is not a supported recovery method, and recovery of logs that were not exported cannot be guaranteed. The recommended next step is to open an Azure Support request for backend investigation and avoid further disruptive changes until the current state has been reviewed.

I hope this helps. Please let us know if you have any further questions.

If this answer helps, please consider clicking Accept Answer so that other community members facing similar issues can benefit from it as well.Hi @megamingus ,

Thank you for reaching out to Microsoft Q&A.

Based on the troubleshooting completed, this appears to be a resource-level diagnostic export issue, rather than a PostgreSQL engine, Log Analytics capacity, or destination-specific problem.

The main indicators are

All diagnostic categories stopped exporting within the same time window.

The PostgreSQL Flexible Server remains healthy and continues generating activity.

Deleting and recreating the diagnostic setting did not restore the export.

Restarting the server had no effect.

A separate diagnostic setting targeting an Azure Storage account also received no data.

The Log Analytics workspace shows no quota, throttling, or ingestion-related events.

Because both Log Analytics and Storage destinations are affected, the failure appears to occur before the logs reach either destination.

1. Can the customer inspect or reset the export pipeline

There is no documented customer-facing mechanism to inspect, restart, or reset the underlying diagnostic export pipeline for an individual PostgreSQL Flexible Server.

Recreating the diagnostic setting is the appropriate customer-side action. Since this has already been completed without restoring the export, Microsoft Support will need to investigate the backend state.

1. Should compute scaling be used as a workaround

A compute scale operation is not a documented or guaranteed remediation for diagnostic export failures. Scaling a production server solely to restore logging is not recommended, particularly before Support has reviewed the existing failure state.

1. Can Microsoft restore the export

Microsoft Support can investigate the resource-level diagnostic routing state and engage the appropriate Azure Database for PostgreSQL and Azure Monitor engineering teams if backend remediation is required.

The support request should include:

Subscription ID and server resource ID

Region: West Europe

Failure window: 24 August 2026, approximately 08:43–08:57 UTC

Diagnostic-setting configuration and resource ID

Log Analytics workspace and Storage account details

Last ingestion time for each affected table

Results of the diagnostic-setting recreation and server restart

Relevant Activity Log, Resource Health, and Service Health results

1. Can the missing logs be recovered

Logs that were not delivered to Log Analytics or Storage should not be assumed to be recoverable or replayable.

As an immediate step, check PostgreSQL Flexible Server → Monitoring → Server logs. PostgreSQL server log files may still be available for download if they remain within the configured retention period. However, this would not necessarily recover other categories such as Query Store Runtime, session statistics, autovacuum statistics, or PgBouncer logs.

For reference:

Configure and access logs in Azure Database for PostgreSQL Flexible Server

Logs in Azure Database for PostgreSQL Flexible Server

Diagnostic settings in Azure Monitor

In summary, the available evidence isolates the failure to the diagnostic export path for this resource. There is no documented customer-side reset, compute scaling is not a supported recovery method, and recovery of logs that were not exported cannot be guaranteed. The recommended next step is to open an Azure Support request for backend investigation and avoid further disruptive changes until the current state has been reviewed.

I hope this helps. Please let us know if you have any further questions.

If this answer helps, please consider clicking Accept Answer so that other community members facing similar issues can benefit from it as well.Thank you for reaching out to Microsoft Q&A.

Based on the troubleshooting completed, this appears to be a resource-level diagnostic export issue, rather than a PostgreSQL engine, Log Analytics capacity, or destination-specific problem.

The main indicators are:

All diagnostic categories stopped exporting within the same time window.

The PostgreSQL Flexible Server remains healthy and continues generating activity.

Deleting and recreating the diagnostic setting did not restore the export.

Restarting the server had no effect.

A separate diagnostic setting targeting an Azure Storage account also received no data.

The Log Analytics workspace shows no quota, throttling, or ingestion-related events.

Because both Log Analytics and Storage destinations are affected, the failure appears to occur before the logs reach either destination.

1. Can the customer inspect or reset the export pipeline

There is no documented customer-facing mechanism to inspect, restart, or reset the underlying diagnostic export pipeline for an individual PostgreSQL Flexible Server.

Recreating the diagnostic setting is the appropriate customer-side action. Since this has already been completed without restoring the export, Microsoft Support will need to investigate the backend state.

1. Should compute scaling be used as a workaround

A compute scale operation is not a documented or guaranteed remediation for diagnostic export failures. Scaling a production server solely to restore logging is not recommended, particularly before Support has reviewed the existing failure state.

1. Can Microsoft restore the export

Microsoft Support can investigate the resource-level diagnostic routing state and engage the appropriate Azure Database for PostgreSQL and Azure Monitor engineering teams if backend remediation is required.

The support request should include:

Subscription ID and server resource ID

Region: West Europe

Failure window: 24 August 2026, approximately 08:43–08:57 UTC

Diagnostic-setting configuration and resource ID

Log Analytics workspace and Storage account details

Last ingestion time for each affected table

Results of the diagnostic-setting recreation and server restart

Relevant Activity Log, Resource Health, and Service Health results

1. Can the missing logs be recovered

Logs that were not delivered to Log Analytics or Storage should not be assumed to be recoverable or replayable.

As an immediate step, check PostgreSQL Flexible Server → Monitoring → Server logs. PostgreSQL server log files may still be available for download if they remain within the configured retention period. However, this would not necessarily recover other categories such as Query Store Runtime, session statistics, autovacuum statistics, or PgBouncer logs.

For reference:

Configure and access logs in Azure Database for PostgreSQL Flexible Server

Logs in Azure Database for PostgreSQL Flexible Server

Diagnostic settings in Azure Monitor

In summary, the available evidence isolates the failure to the diagnostic export path for this resource. There is no documented customer-side reset, compute scaling is not a supported recovery method, and recovery of logs that were not exported cannot be guaranteed. The recommended next step is to open an Azure Support request for backend investigation and avoid further disruptive changes until the current state has been reviewed.

I hope this helps. Please let us know if you have any further questions.

If this answer helps, please consider clicking Accept Answer so that other community members facing similar issues can benefit from it as well.

Hello @megamingus

Given the results of your previous tests, particularly the fact that both the Log Analytics and the separate Storage Account destination have stopped receiving the logs, I agree that this should now be investigated by Microsoft Support.

To answer your questions directly:

1. Is there a method available for checking or resetting the diagnostic export pipeline?

Nothing can be done from the customer's side apart from managing or recreating the diagnostic settings and destinations. There is no documented procedure to restart the underlying PostgreSQL/Azure Monitor diagnostic export pipeline; since recreating the diagnostic setting had no effect, Microsoft Support would need to investigate the backend state.

2. Would increasing the size of the PostgreSQL server cause the pipeline to recover?

It may be a side effect of moving or reconfiguring the underlying computing resources, but this is not a documented or guaranteed way to fix the problem by exporting diagnostic logs. There's no point in scaling the production server for this, especially before Support has had a chance to review the current failure state.

3. Can Microsoft fix the diagnostic pipeline?

If the issue lies with the service-side export pipeline, Microsoft would need to investigate and fix it. The test you ran using your Storage Account is useful evidence, as it indicates the problem occurs before the destination and not specifically with Log Analytics.

4. Is it possible to recover the missing logs since about 08:57 UTC?

It would be incorrect to assume they can be replayed. Diagnostic settings stream the logs to the specified destination and are not meant to function as a separate store from which customers can access the data. Nevertheless, you should look at the PostgreSQL Flexible Server → Monitoring → Server logs right away. If server logging was enabled and the relevant files are still within the retention period, you may be able to retrieve some PostgreSQL logs without relying on the Azure Monitor diagnostic export.

Then open the support case and provide Microsoft with the server resource ID, the West Europe region, the failure timestamp (between 08:43 and 08:57 UTC), the diagnostic-setting configuration, and the results of both the Log Analytics and Storage Account tests.

You should also put off any further disruptive changes, such as scaling or redeployment, until Support has taken control of the backend state.

Microsoft documentation on PostgreSQL Flexible Server logging and diagnostic settings:

Configure and access logs for PostgreSQL Flexible Server

Help make this community better for everyone: if this answer resolved your issue, please accept it or leave an upvote. If not, share more details in a comment so we can continue the discussion and find the right solution.