Azure Comms Service returning RestError due to time difference when email volume is high

Question

Azure Comms Service returning RestError due to time difference when email volume is high

Matt Worsley 20

I'm using the @azure/communication-email EmailClient SDK in my Node server (hosted as an App Service). The client listens for 'email-ready' events my Service Bus (via business logic in my app) and sends the email using emailClient.beginSend().

It works fine apart from when volumes are higher, say 100 sends per minute. In those cases it returns, intermittently, the error :

RestError: The given request could not be resolved. The time difference between the originating client and the server is greater than the allowed margin of 5 minutes. at handleErrorResponse (C:\home\site\wwwroot\node_modules\@azure\core-client\dist\commonjs\deserializationPolicy.js:148:19) at deserializeResponseBody (C:\home\site\wwwroot\node_modules\@azure\core-client\dist\commonjs\deserializationPolicy.js:83:45) at process.processTicksAndRejections (node:internal/process/task_queues:95:5)

I can't find any guidance on this error in the docs. All I can do at the moment is implement retry logic, but if the volume is consitantly high, it just keeps raising the same error over and over.

I have checked on Kudu and the server time on my App Service is correct and hasn't drifted during these peak periods

Siva Nair 2,420 Reputation points Microsoft External Staff Moderator

2025-03-17T12:09:47.57+00:00
Hi Matt Worsley,

Please refer the similar cases which may help you to resolve ,

https://learn.microsoft.com/en-us/answers/questions/1820011/time-service-on-server

https://learn.microsoft.com/en-us/answers/questions/45406/changing-date-and-time-on-azure-web-app

https://learn.microsoft.com/en-us/azure/virtual-machines/windows/time-sync

Moreover ,

Verify NTP Configuration: Ensure your server synchronizes with a reliable Network Time Protocol (NTP) server. Misconfigurations can lead to time discrepancies.

Monitor Time Drift: Regularly monitor your server's time, especially under high load, to detect and correct any drift promptly. Tools like w32tm (for Windows) can help check the status and configuration of the time service.

Assess Network Latency: Monitor network performance to ensure that delays aren't contributing to perceived time differences.

Implement Efficient Retry Logic: While you've added retry mechanisms, ensure they're optimized to handle transient issues without overwhelming the system.

Throttle Requests: Implement rate limiting to prevent overwhelming the ACS with high volumes of requests in a short period.

Batch Processing: Consider batching emails to manage load more effectively and reduce the chance of time-related errors.

If you have any further assistant, do let me know.
Siva Nair 2,420 Reputation points Microsoft External Staff Moderator

2025-03-18T01:14:02.4466667+00:00

Hi Matt Worsley,

We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.
Siva Nair 2,420 Reputation points Microsoft External Staff Moderator

2025-03-18T16:43:52.82+00:00

Hi Matt Worsley,

We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.
Siva Nair 2,420 Reputation points Microsoft External Staff Moderator

2025-03-19T08:17:51.36+00:00

Hi Matt Worsley,

We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.

Matt Worsley 20

Hi Siva - The app is hosted by Azure, so presume the time configs are all ok. I don't even think I have access to them. The retry mechanisms I have are adequate, but do not solve the problem where there are high priority emails floating around the retry mechanism for 30 minutes. Regarding batching, I can't see any API for sending batch emails to Azure Communications Service, only single requests.

I have also tried spawning new EmailClient instances when there are more than 5 sends being currently handled by the current instance, but it hasn't resolved it. This is my current send operation, if it helps..

      const date = new Date().toUTCString();

      const poller = await this.emailClient.beginSend(message, {
        requestOptions: {
          //an attempt to stop "RestError: The given request could not be resolved. The time difference between the originating client and the server is greater than the allowed margin of 5 minutes."
          customHeaders: { "x-ms-date": date },
          timeout: one("minute"),
        },
      });
      this.inProgress++;

      if (!poller.getOperationState().isStarted) {
        throw new Error(
          `Failed to send the email (operation id: ${poller.getResult().id})`
        );
      }

      await poller.pollUntilDone();
      this.inProgress--;
      const info = {
        message: `Successfully sent the email (operation id: ${
          poller.getResult().id
        })`,
        result: poller.getResult(),
      };
      return info;

Accepted answer

0 additional answers

Your answer

Siva Nair 2,420 Reputation points Microsoft External Staff Moderator

2025-03-17T12:09:47.57+00:00

Hi Matt Worsley,

Please refer the similar cases which may help you to resolve ,

https://learn.microsoft.com/en-us/answers/questions/1820011/time-service-on-server

https://learn.microsoft.com/en-us/answers/questions/45406/changing-date-and-time-on-azure-web-app

https://learn.microsoft.com/en-us/azure/virtual-machines/windows/time-sync

Moreover ,

Verify NTP Configuration: Ensure your server synchronizes with a reliable Network Time Protocol (NTP) server. Misconfigurations can lead to time discrepancies.

Monitor Time Drift: Regularly monitor your server's time, especially under high load, to detect and correct any drift promptly. Tools like w32tm (for Windows) can help check the status and configuration of the time service.

Assess Network Latency: Monitor network performance to ensure that delays aren't contributing to perceived time differences.

Implement Efficient Retry Logic: While you've added retry mechanisms, ensure they're optimized to handle transient issues without overwhelming the system.

Throttle Requests: Implement rate limiting to prevent overwhelming the ACS with high volumes of requests in a short period.

Batch Processing: Consider batching emails to manage load more effectively and reduce the chance of time-related errors.

If you have any further assistant, do let me know.
Siva Nair 2,420 Reputation points Microsoft External Staff Moderator

2025-03-18T01:14:02.4466667+00:00

Hi Matt Worsley,

We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.
Siva Nair 2,420 Reputation points Microsoft External Staff Moderator

2025-03-18T16:43:52.82+00:00

Hi Matt Worsley,

We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.
Siva Nair 2,420 Reputation points Microsoft External Staff Moderator

2025-03-19T08:17:51.36+00:00

Hi Matt Worsley,

We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution, please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.
Matt Worsley 20 Reputation points

2025-03-26T15:31:17.9133333+00:00

Hi Siva - The app is hosted by Azure, so presume the time configs are all ok. I don't even think I have access to them. The retry mechanisms I have are adequate, but do not solve the problem where there are high priority emails floating around the retry mechanism for 30 minutes. Regarding batching, I can't see any API for sending batch emails to Azure Communications Service, only single requests.

I have also tried spawning new EmailClient instances when there are more than 5 sends being currently handled by the current instance, but it hasn't resolved it. This is my current send operation, if it helps..

const date = new Date().toUTCString(); const poller = await this.emailClient.beginSend(message, { requestOptions: { //an attempt to stop "RestError: The given request could not be resolved. The time difference between the originating client and the server is greater than the allowed margin of 5 minutes." customHeaders: { "x-ms-date": date }, timeout: one("minute"), }, }); this.inProgress++; if (!poller.getOperationState().isStarted) { throw new Error( `Failed to send the email (operation id: ${poller.getResult().id})` ); } await poller.pollUntilDone(); this.inProgress--; const info = { message: `Successfully sent the email (operation id: ${ poller.getResult().id })`, result: poller.getResult(), }; return info;

Answer 1

Hi Matt Worsley,

ACS uses shared access keys under the hood to sign each request with a timestamp, and the backend checks that the time difference is within a 5-minute window. If your EmailClient instance is long-lived, or if there’s a delay between token creation and request execution (like retries or queue delays), that timestamp can go stale — and ACS rejects it.

Rather than reusing a single client instance across all sends, I now just create a new one per email (or per small batch). That ensures the timestamp used for signing is always fresh and reduces the chance of drift.

const emailClient = new EmailClient(process.env.EMAIL_CONNECTION_STRING);
const poller = await emailClient.beginSend(message);
await poller.pollUntilDone();

If you’re in a position to use Azure AD (via DefaultAzureCredential or ClientSecretCredential), do it. These tokens are refreshed automatically and are much more resilient under load. You’ll avoid the time skew problem entirely.

const credential = new DefaultAzureCredential();
const emailClient = new EmailClient(endpoint, credential);

Unfortunately, ACS doesn’t support batch email sends yet — each beginSend() call is for one email. So you’ve got to optimize around single sends.

Hope it helps!

Please do not forget to click "Accept the answer” and Yes wherever the information provided helps you, this can be beneficial to other community members.

User's image

If you have any other questions or still running into more issues, let me know in the "comments" and I would be happy to help you.

Suresh Chikkam 2,135 Reputation points Microsoft External Staff Moderator

2025-04-04T06:21:48.9766667+00:00

Hi Matt Worsley,

Following up to see if the above answer was helpful. If this answers your query, do click Accept Answer and Yes, if you have any further query do let us know.

Share via

Azure Comms Service returning RestError due to time difference when email volume is high

0 additional answers

Your answer