Flex Consumption: storage queue trigger never scaled out from zero for 4 days despite continuous queue backlog

Question

Flex Consumption: storage queue trigger never scaled out from zero for 4 days despite continuous queue backlog

동윤 이 20

Environment

Azure Functions Flex Consumption plan, Korea Central
Python 4.x host (host version 4.1052.300.26370), single storage queue trigger function in the app
host.json queues settings: batchSize 16, maxDequeueCount 5, visibilityTimeout 00:01:00

Symptom

Our function app has a single Azure Storage Queue trigger function. After the app scaled in to zero instances on 2026-08-10 07:07 UTC, the platform scale controller did not start any instance for the queue function group for about 4 days, even though messages were continuously enqueued to the monitored queue throughout the entire window. Processing only resumed on 2026-08-14 05:24:36 UTC, when instances with "Function group target is function:

Pravallika KV 18,855 평판 포인트 Microsoft 외부 직원 중재자

2026-08-23T19:56:35.67+00:00

@동윤 이 , 개인 메시지로 연락드렸습니다. 요청드린 정보를 공유해 주시기 바랍니다.
Sridevi Machavarapu 33,820 평판 포인트 Microsoft 외부 직원 중재자

2026-08-30T04:46:00.98+00:00

Hello 동윤 이,

We have completed the investigation and identified a platform-side issue in the Azure Functions host that caused the Storage Queue trigger configuration to be published incorrectly during host shutdown. This caused the queue to be reported as empty even though messages were pending, so the app did not scale out from zero.

Processing resumed on August 14 after the trigger configuration was synchronized. We also confirmed that no deployment or configuration changes were made from your side during the affected period.

If the issue occurs again, syncfunctiontriggers can be used to restore the trigger configuration. For this production workload, engineering also recommends Always Ready as a preventive measure.

The underlying platform issue is being tracked by the engineering team.

Please let us know if you have any further questions.
Sridevi Machavarapu 33,820 평판 포인트 Microsoft 외부 직원 중재자

2026-08-31T01:33:51.3466667+00:00

Hello 동윤 이,

Please confirm whether you still have further queries on this or issue resolved.

질문 작성자가 수락한 답변

0 추가 답변

답변

Pravallika KV 18,855 평판 포인트 Microsoft 외부 직원 중재자

2026-08-23T19:56:35.67+00:00

@동윤 이 , 개인 메시지로 연락드렸습니다. 요청드린 정보를 공유해 주시기 바랍니다.
Sridevi Machavarapu 33,820 평판 포인트 Microsoft 외부 직원 중재자

2026-08-30T04:46:00.98+00:00

Hello 동윤 이,

We have completed the investigation and identified a platform-side issue in the Azure Functions host that caused the Storage Queue trigger configuration to be published incorrectly during host shutdown. This caused the queue to be reported as empty even though messages were pending, so the app did not scale out from zero.

Processing resumed on August 14 after the trigger configuration was synchronized. We also confirmed that no deployment or configuration changes were made from your side during the affected period.

If the issue occurs again, syncfunctiontriggers can be used to restore the trigger configuration. For this production workload, engineering also recommends Always Ready as a preventive measure.

The underlying platform issue is being tracked by the engineering team.

Please let us know if you have any further questions.
Sridevi Machavarapu 33,820 평판 포인트 Microsoft 외부 직원 중재자

2026-08-31T01:33:51.3466667+00:00

Hello 동윤 이,

Please confirm whether you still have further queries on this or issue resolved.

Answer 1

Sridevi Machavarapu 33,820 Microsoft 외부 직원 중재자

Hello 동윤 이,

We have completed the investigation and identified a platform-side issue in the Azure Functions host that caused the Storage Queue trigger configuration to be published incorrectly during host shutdown. As a result, the queue was reported as empty even though messages were pending, so the app did not scale out from zero.

Processing resumed on August 14 after the trigger configuration was synchronized. We also confirmed that no deployment or configuration changes were made from your side during the affected period.

If the issue occurs again, syncfunctiontriggers can be used to restore the trigger configuration. For this production workload, engineering also recommends Always Ready as a preventive measure.

The underlying platform issue is being tracked by the engineering team.

Please let us know if you have any further questions.

If this answers your query, do click Accept Answer and Yes for was this answer helpful, which may help members with similar questions.

동윤 이 20 평판 포인트

2026-09-01T06:56:40.77+00:00

Hello @Sridevi Machavarapu ,

Thank you for the detailed explanation. I understand the root cause you described and the recommended steps, and I have no further questions at this point.

I appreciate you and Pravallika looking into this, and I understand the underlying platform issue is being tracked by the engineering team.
Sridevi Machavarapu 33,820 평판 포인트 Microsoft 외부 직원 중재자

2026-09-02T11:55:56.6533333+00:00

Hello 동윤 이,

Thank you for confirming. We appreciate your patience and cooperation while we investigated the issue.

If this answers your query, please click Accept Answer and Yes for Was this answer helpful? This may also help other community members facing a similar issue.