Things I've learned about sending email from IBM i apps

I spent some time this past week, vetting and validating how the IBM i SMTP
client works; how it might be used for sending email from IBM i apps.
Following are some key points.

The SMTP client offers at least 4 ways for sending email:

1. Sending directly to any email server on the planet that has a DNS MX
record defined for a service on port 25, which essentially means being able
to send to anyone who has an email account anywhere.

2. Forwarding mail to a configured mail hub, by using credentials (account
and password) that are authenticated by the mail hub. The mail hub then
forwards to the final destination point, using the sender's account which
was provided by the IBM i SMTP client.

3. Forwarding mail to a configured relay-hub, that may or may not require
credentials (relay hubs may be configured to allow relay from specific IP
address ranges). Relay hubs may offer quite a few options for handling
email.

4. Forwarding mail to IBM i users, using IBM i distribution services (i.e.
mail between IBM i systems).

The IBM i SMTP client can be configured to support SSL and TLS encryption,
which may or may not be required by a mail or relay hubs or final end
point, which is negotiable when the connection is made.

I tested forwarding mail through Google mail hubs, using methods 2 and 3.

The most challenging part to using a mail hub or relay service is importing
the right certificates into IBM i certificate manager and configuring the
IBM i SMTP client to use them.

If you run into problems interfacing with a mail hub or relay service, the
TRCTCPAPP command can be used to view the conversation that occurs under
the covers.

When sending mail directly to end points, configuring DNS "spf" and "dmarc"
entries can make a world of difference in whether email gets blocked or
routed to spam or not.

HTH,

Nathan.