Connecting to external SMTP server from Web App

Question

Connecting to external SMTP server from Web App

Nick Roberts 6

I'm got an Azure Web App running and one of it's tasks is to send emails out in response to user requests.

It does this by connecting to our (external) WHM/cPanel-based mail server using an authenticated, secure connection on port 465.

The Web App is running node.js 12.x and I'm using the nodemailer library.

Up until 1st April 2021 this was running fine, now not. When I look at the logs, I can see the DNS lookup working fine, but the connection is never established:

DEBUG Sending mail using SMTP/6.4.11[client:6.4.11]
DEBUG [LFBZLRxTUro] Resolved xxx.xxx.com as x.x.x.x [cache miss]

The same code running on my local machine runs fine:

DEBUG Sending mail using SMTP/6.4.11[client:6.4.11]
DEBUG [6rjf9wZwHAY] Resolved xxx.xxx.com as x.x.x.x [cache miss]
INFO  [6rjf9wZwHAY] Secure connection established to x.x.x.x:465

Two questions:

Is what I'm trying to do possible on Azure using the Web Apps service
Why was this AOK before 1st April 2021, and now bust?

SnehaAgrawal-MSFT 22,711 Reputation points Moderator

2021-04-07T05:56:00.337+00:00

Thanks for asking question! Could you please take a look if there is any firewall issue with your SMTP server.

Also, you may want to know that Azure, you don't need to use External SMTP server to send emails in your app and you can easily use SendGrid to satisfy this.

You may refer to detailed guide on how to use the SendGrid email service from PHP: https://azure.microsoft.com/en-us/documentation/articles/store-sendgrid-php-how-to-send-email/.

Also check this SO link might help.

Let us know if further query or issue remains.
Nick Roberts 6 Reputation points

2021-04-07T08:32:30.96+00:00

Thanks for the note.

I've checked my email servers firewall settings and everything is fine there - port 465 is open to TCP connections from all IPs.

This stopped working on 1st April - the first failure I have in the logs is at 1852 UTC. There were two resource health issues that day at 0357 and 0842 UTC.

These were the only changes to any of these systems that day - so I'm just more than a bit suspicious that this was caused by changes at the Azure end.

I appreciate the SendGrid suggestion, but this is not what I want to do here - I want the code base I've written (and been running for nearly a year) to continue to work using my SMTP server.
Nick Roberts 6 Reputation points

2021-04-07T10:07:33.653+00:00

Further details from MemSet (who operate the SMTP server I use):

Hi Nicholas,

Thanks for contacting Memset Support today.

I can confirm that Memset do not block any Azure/Microsoft IP Addresses. I can also confirm that Memset does not make any changes to customer servers prior to asking permission or letting them know.
Nick Roberts 6 Reputation points

2021-04-07T15:29:58.227+00:00

Looking more closely into our logs, all from 1st April

Email send attempt at 02:06:44 UTC [success]
Email send attempt at 13:11:47 UTC [failure]
Email send attempt at 18:52:46UTC [failure]

All others after this fail.

Note that this server sends very little email (maybe 10/day max), hence the 11-hour gap between the first two sends.

Your answer

SnehaAgrawal-MSFT 22,711 Reputation points Moderator

2021-04-07T05:56:00.337+00:00

Thanks for asking question! Could you please take a look if there is any firewall issue with your SMTP server.

Also, you may want to know that Azure, you don't need to use External SMTP server to send emails in your app and you can easily use SendGrid to satisfy this.

You may refer to detailed guide on how to use the SendGrid email service from PHP: https://azure.microsoft.com/en-us/documentation/articles/store-sendgrid-php-how-to-send-email/.

Also check this SO link might help.

Let us know if further query or issue remains.
Nick Roberts 6 Reputation points

2021-04-07T08:32:30.96+00:00

Thanks for the note.

I've checked my email servers firewall settings and everything is fine there - port 465 is open to TCP connections from all IPs.

This stopped working on 1st April - the first failure I have in the logs is at 1852 UTC. There were two resource health issues that day at 0357 and 0842 UTC.

These were the only changes to any of these systems that day - so I'm just more than a bit suspicious that this was caused by changes at the Azure end.

I appreciate the SendGrid suggestion, but this is not what I want to do here - I want the code base I've written (and been running for nearly a year) to continue to work using my SMTP server.
Nick Roberts 6 Reputation points

2021-04-07T10:07:33.653+00:00

Further details from MemSet (who operate the SMTP server I use):

Hi Nicholas,

Thanks for contacting Memset Support today.

I can confirm that Memset do not block any Azure/Microsoft IP Addresses. I can also confirm that Memset does not make any changes to customer servers prior to asking permission or letting them know.
Nick Roberts 6 Reputation points

2021-04-07T15:29:58.227+00:00

Looking more closely into our logs, all from 1st April

Email send attempt at 02:06:44 UTC [success]
Email send attempt at 13:11:47 UTC [failure]
Email send attempt at 18:52:46UTC [failure]

All others after this fail.

Note that this server sends very little email (maybe 10/day max), hence the 11-hour gap between the first two sends.

Share via

Connecting to external SMTP server from Web App

Your answer