An API that connects multiple Microsoft services, enabling data access and automation across platforms
Thank you for the detailed explanation of the asynchronous sendMail pipeline.
Based on your response, we understand that:
- The sendMail operation returns
202 Acceptedafter draft creation - Further processing occurs within Exchange Online transport
- There is no explicit idempotency guarantee at the service level
However, we would like to highlight the following observations from our system:
Key Findings:
- Only a single API request is made per execution (confirmed via server logs)
- Only one
request-idis generated per send operation - No retry logic exists in our application
- Only one instance of the service is running (PM2 single process)
- Duplicate emails are received with identical content and near-identical timestamps
- This issue occurs intermittently (not on every execution)
Example Case:
- Request ID:
8abb86da-90a1-4637-9471-f99ee8604b2f - Timestamp (UTC):
Wed, 06 May 2026 05:40:06 GMT - Endpoint:
POST /v1.0/users/{user}/sendMail
Clarification Required: Given that:
- Only one request is issued
- No client-side retries occur
We would like to confirm:
- Is it possible for Exchange Online transport to internally reprocess the same message and deliver it more than once?
- Are there known scenarios where a single accepted request may result in duplicate delivery without generating an NDR?
- Can you verify, using the above request-id, whether the message was processed more than once within the transport pipeline?
Request: We kindly request an internal trace analysis using the provided request-id to confirm:
- Number of times the message entered the transport pipeline
- Whether any internal retry or duplication occurred
As this behavior impacts production reliability, we need to determine whether additional safeguards must be implemented on our side or if this is an issue within the service pipeline.
We are happy to provide additional logs or correlation data if required.
Thank you for your support.Thank you for the detailed explanation of the asynchronous sendMail pipeline.
Based on your response, we understand that:
- The sendMail operation returns
202 Acceptedafter draft creation - Further processing occurs within Exchange Online transport
- There is no explicit idempotency guarantee at the service level
However, we would like to highlight the following observations from our system:
Key Findings:
- Only a single API request is made per execution (confirmed via server logs)
- Only one
request-idis generated per send operation - No retry logic exists in our application
- Only one instance of the service is running (PM2 single process)
- Duplicate emails are received with identical content and near-identical timestamps
- This issue occurs intermittently (not on every execution)
Example Case:
- Request ID:
8abb86da-90a1-4637-9471-f99ee8604b2f - Timestamp (UTC):
Wed, 06 May 2026 05:40:06 GMT - Endpoint:
POST /v1.0/users/{user}/sendMail
Clarification Required:
Given that:
- Only one request is issued
- No client-side retries occur
We would like to confirm:
- Is it possible for Exchange Online transport to internally reprocess the same message and deliver it more than once?
- Are there known scenarios where a single accepted request may result in duplicate delivery without generating an NDR?
- Can you verify, using the above request-id, whether the message was processed more than once within the transport pipeline?
Request:
We kindly request an internal trace analysis using the provided request-id to confirm:
- Number of times the message entered the transport pipeline
- Whether any internal retry or duplication occurred
As this behavior impacts production reliability, we need to determine whether additional safeguards must be implemented on our side or if this is an issue within the service pipeline.
We are happy to provide additional logs or correlation data if required.
Thank you for your support.