• Check whether the stuck mails have something in common. (size/attachments/receipients...)
• Search for errors in sogo.log around: ... POST /SOGo/Microsoft-Server-ActiveSync?Cmd=SendMail&DeviceType=....

We are having this issue with all ActiveSync clients.

Sendmail returns “user unknown”, but the ActiveSync client keeps retrying.

e.g

Oct 19 11:13:29 mail sendmail[1278]: v9J9DT6i001278: <user@example.com>... User unknown
Oct 19 11:25:37 mail sendmail[4673]: v9J9Pbod004673: <user-at-example.com>... User unknown

This happens when sending to a non-existent mailbox hosted locally and any invalid email address.

To resolve this we create a temporary alias to an existing mailbox or ask the user to remove
The invalid email address from the email.

I have same problem. That's for sure bug in SOGo - when mail is sent (via webmail or EAS) it is passed to SMTP server, which can accept or reject it depending on configuration. For example Postfix has option smtpd_recipient_restrictions=reject_unknown_recipient_domain which rejects inbound SMTP from SOGo when non-existent recipient domain in To: field. When SOGo webmail don't even allow to send this message (saying invalid recipient), EAS implementation ignores SMTP rejection and mail sits in Outbox forever and, depending on client, delivery attempts continue.

To work around this, you can create separate submission service without any checks. Heres excerpt from my Postfix master.cf:

# wmsub_relay
5877       inet n       -       y       -       -       smtpd
  -o syslog_name=postfix/wmsub_relay
  -o smtpd_sasl_auth_enable=no
  -o smtpd_reject_unlisted_recipient=no
  -o smtpd_client_restrictions=
  -o smtpd_helo_restrictions=
  -o smtpd_sender_restrictions=
  -o smtpd_recipient_restrictions=
  -o mynetworks=127.0.0.1/32
  -o smtpd_relay_restrictions=permit_mynetworks,reject
  -o milter_macro_daemon_name=ORIGINATING

This way instead of rejecting, MTA accepts mail (after that it disapears from EAS outbox) and then after delivery attempt sends non-delivery notification back to sender.

But I hope this issue will be fixed in SOGo in future.

I can confirm this problem as well.

If you send an eMail to an internal user, who does not exist.

Log:
"sogod[67:67] SMTP(RCPT TO) error: 5.1.1 <destination@mail.mail>: Recipient address rejected: User unknown in virtual mailbox table

How can i support, that this problem can be fixed?

this problems only exists, if you send to an internal not-existant email address. if it is an external not-existant email address, the email gets out....

workaround is to disable smtpd_reject_unlisted_recipient in postfix.

with:

 -o smtpd_reject_unlisted_recipient=no # SoGo EAS Fix

We are facing the same problem. However I've noticed one strange thing in logs:

"User unknown" in postfix  --->> "Could not connect to the SMTP server" in SOGO -->> Outlook

Jun 30 12:02:34 mail postfix/submission/smtpd[399412]: connect from localhost[127.0.0.1]
Jun 30 12:02:34 mail postfix/submission/smtpd[399412]: Anonymous TLS connection established from localhost[127.0.0.1]: TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256
Jun 30 12:02:34 mail postfix/submission/smtpd[399412]: NOQUEUE: reject: RCPT from localhost[127.0.0.1]: 550 5.1.1 <sdfdjfdshj@zim-service.xxx>: Recipient address rejected: User unknown; from=<support@zim-service.xxx> to=<sdfdjfdshj@zim-ser
vice.xxx> proto=ESMTP helo=<localhost>
Jun 30 12:02:34 mail postfix/submission/smtpd[399412]: lost connection after RCPT from localhost[127.0.0.1]
Jun 30 12:02:34 mail postfix/submission/smtpd[399412]: disconnect from localhost[127.0.0.1] ehlo=2 starttls=1 auth=1 mail=1 rcpt=0/1 commands=5/6

2022-06-30 12:02:34.874 sogod[201406:201406] SMTP: STARTTLS successfully performed 
2022-06-30 12:02:34.898 sogod[201406:201406] SMTP(RCPT TO) error: 5.1.1 sdfdjfdshj@zim-service.xxx: Recipient address rejected: User unknown 
Jun 30 12:02:34 sogod [201406]: [ERROR] <0x0x564794b995c0[SOGoMailer]> Could not connect to the SMTP server smtp://127.0.0.1:587/?tls=YES&tlsVerifyMode=allowInsecureLocalhost 
Jun 30 12:02:34 sogod [201406]: localhost "POST /SOGo/Microsoft-Server-ActiveSync?Cmd=SendMail&User=support@zim-service.xxx&DeviceId=76D251AB1CDC4CC7801AD175138AF169&DeviceType=WindowsOutlook15 HTTP/1.0" 200 15/2631 0.039 - - 0 - 25 
Jun 30 12:02:35 sogod [1144]: [WARN] <0x0x56479361ba10[WOWatchDogChild]> pid 301948 has been hanging in the same request for 10 minutes 
Jun 30 12:02:35 sogod [1144]: [WARN] <0x0x56479363ae70[WOWatchDogChild]> pid 371783 has been hanging in the same request for 55 minutes

Now that new guys have been taking over this project. Can you please look into this again?