TidSmtpServer hangs on some connections

[bluewwol]

Hi,

I am hoping for some advice on how to proceed to resolve the following incoming mail issue.

Ordinarily transactions come in like this and are handled perfectly

Code:

70.xxx.xxx.xxx:55906 Stat Connected.
70.xxx.xxx.xxx:55906 Sent 1/6/2019 6:36:45 PM: 220 Welcome to WWOL Server<EOL>
70.xxx.xxx.xxx:55906 Recv 1/6/2019 6:36:45 PM: EHLO mx01.kmtel.com<EOL>
70.xxx.xxx.xxx:55906 Sent 1/6/2019 6:36:45 PM: 250-Hello mx01.kmtel.com<EOL>250-AUTH LOGIN<EOL>250-ENHANCEDSTATUSCODES<EOL>250-PIPELINING<EOL>250-SIZE 0<EOL>250-CHUNKING<EOL>250-8BITMIME<EOL>250 BINARYMIME<EOL>
70.xxx.xxx.xxx:55906 Recv 1/6/2019 6:36:45 PM: MAIL FROM:<MAILER-DAEMON@kmtel.com> SIZE=10078<EOL>
70.xxx.xxx.xxx:55906 Sent 1/6/2019 6:36:45 PM: 250 2.1.0 MAILER-DAEMON@kmtel.com Address Okay<EOL>
70.xxx.xxx.xxx:55906 Recv 1/6/2019 6:36:45 PM: RCPT TO:<info@domain.com><EOL>
70.xxx.xxx.xxx:55906 Sent 1/6/2019 6:36:45 PM: 250 2.1.5 info@domain.com Address Okay<EOL>
70.xxx.xxx.xxx:55906 Recv 1/6/2019 6:36:46 PM: DATA<EOL>
70.xxx.xxx.xxx:55906 Sent 1/6/2019 6:36:46 PM: 354 Start mail input; end with <CRLF>.<CRLF><EOL>

however for some the negotiation looks like this

Code:

52.xxx.xxx.xxx:60351 Stat Connected.
52.xxx.xxx.xxx:60351 Sent 1/6/2019 6:34:44 PM: 220 Welcome to WWOL Server<EOL>
52.xxx.xxx.xxx:60351 Recv 1/6/2019 6:34:44 PM: EHLO NAM01-BY2-obe.outbound.protection.outlook.com<EOL>
52.xxx.xxx.xxx:60351 Sent 1/6/2019 6:34:44 PM: 250-Hello NAM01-BY2-obe.outbound.protection.outlook.com<EOL>250-AUTH LOGIN<EOL>250-ENHANCEDSTATUSCODES<EOL>250-PIPELINING<EOL>250-SIZE 0<EOL>250-CHUNKING<EOL>250-8BITMIME<EOL>250 BINARYMIME<EOL>
52.xxx.xxx.xxx:60351 Recv 1/6/2019 6:34:44 PM: MAIL FROM:<> SIZE=83225<EOL>RCPT TO:<info@domain.com><EOL>
52.xxx.xxx.xxx:60351 Stat Disconnected.

the difference here is the remote server is compounding the mail from and rcpt to transactions into a single transaction and when this happens Indy never responds to the transaction and the thread hangs until it is cleaned up several minutes later.
Unfortunately several domains including the microsoft domains do this and for me this constitutes a very large number of messages.  
I am hoping someone can advise a way to work around this issue.

Thanks for any guidance,
-Allen

---

I experienced the issue above with PIPELINING set to true.  Changing that to false causes everything to function as expected.  

I guess that answers my question, sorry for the silly question, unless there is something I ought to have done to enable pipeling to function properly.

-Allen