Re: [exim] stuck exim processes

I finally got something I think is useful.
The debug delivery command
exim -d+all -M 1nJzUm-00007r-4z
tries to send out the message but itself is stuck.

Output (after lots of other stuff) looks like this:

07:45:33.757 57451 sending small initial BDAT; hsize=1451
07:45:33.757 57451 SMTP+> BDAT 1451
07:45:33.757 57451 cmd buf flush 101 bytes (more expected)
07:45:33.757 57451 tls_write(0x8011111a0, 101, more)
07:45:33.757 57451 writing data block fd=8 size=1451 timeout=300
07:45:33.757 57451 tls_write(0x8010df808, 1451)
07:45:33.757 57451 SSL_write(0x801600000, 0x8010500d8, 1552)
07:45:33.757 57451 outbytes=1552 error=0
07:45:33.757 57451 look for 2 responses for previous pipelined cmds
07:45:33.757 57451 sync_responses expect mail
07:45:34.638 57451 Calling SSL_read(0x801600000, 0x8011101a0, 4096)
07:45:34.638 57451 read response data: size=147
07:45:34.638 57451 SMTP<< 250 2.1.0 Ok
07:45:34.638 57451 sync_responses expect rcpt for
Recipient@???
07:45:34.638 57451 SMTP<< 550 5.1.1
<xn--recipient-wob@???>: Recipient address rejected:
User unknown
07:45:34.638 57451 look for one response for BDAT
07:45:34.638 57451 SMTP<< 554 5.5.1 Error: no valid recipients
07:45:34.638 57451 ok=0 send_quit=1 send_rset=1 continue_more=0 yield=0
first_address is NULL
07:45:34.638 57451 SMTP+> QUIT
07:45:34.638 57451 cmd buf flush 6 bytes (more expected)
07:45:34.638 57451 tls_write(0x8011111a0, 6, more)
07:45:34.638 57451 tls_write((nil), 0)
07:45:34.638 57451 SSL_write(0x801600000, 0x801050768, 6)
07:45:34.638 57451 outbytes=6 error=0
07:45:34.638 57451 SMTP(TLS shutdown)>>
07:45:34.638 57451 SSL3 alert write:warning:close notify
07:45:34.638 57451 SMTP(shutdown)>>
07:45:34.643 57451 Calling SSL_read(0x801600000, 0x8011101a0, 4096)
07:45:34.643 57451 read response data: size=15
07:45:34.643 57451 SMTP<< 221 2.0.0 Bye
07:45:34.643 57451 Calling SSL_read(0x801600000, 0x8011101a0, 4096)
07:45:35.700 57451 tls_close(): shutting down TLS (with response-wait)
07:45:35.700 57451 tls_write((nil), 0)
07:46:33.400 57448 polling subprocess pipes
07:47:33.420 57448 polling subprocess pipes
07:48:33.450 57448 polling subprocess pipes
07:49:33.520 57448 polling subprocess pipes
07:50:33.530 57448 polling subprocess pipes
07:51:33.550 57448 polling subprocess pipes
07:52:33.650 57448 polling subprocess pipes
07:53:33.730 57448 polling subprocess pipes
07:54:33.750 57448 polling subprocess pipes
07:55:33.830 57448 polling subprocess pipes
07:56:33.930 57448 polling subprocess pipes
07:57:33.950 57448 polling subprocess pipes

Am 14.02.22 um 16:25 schrieb Jeremy Harris via Exim-users:
> On 10/02/2022 15:24, Martin Waschbüsch via Exim-users wrote:
>> If I interpret this correctly, the message was delivered and then the
>> process got stuck. Which is why no additional action but cleanup was
>> needed?
>
> I think it says that a journal file was found, which said that the
> apparently outsnding address had actually been delivered, so
> it tidied up and exited - ie. this process did not get stuck.
> It says explicitly at the end that it is terminating normally.
>
> The journal file, however, is consistent with the probability
> of a crash of a previous transport process.
>


Best,

Martin