Hello Justin and developers

My list has been running reliably since November 2019, with an update to 11.11.2 in September. However, the list has not sent out any messages since 21st May (apart from a message to me on 25th May saying a non-subscriber had tried to post).

The shared host error_log file has messages every five minutes that include "AH01220: Timeout waiting for output from CGI script/home/very_long_path/cgi-bin/dada/mail.cgi" and "Script timed out before returning headers: mail.cgi". My attention was drawn to these by the host support team, who were responding to our webmaster's queries about why our website, on the same host, was slow and giving 504 timeout errors.

When I tried to run the Dada Mail scheduled job from the command line(following the online documentation), mail.cgi took two or three minutes to respond, which explains the timeouts.  I have attached a redacted version of the verbose response eventually obtained. Much of it looks sensible. It also shows that the daily digests that are sent to me have not been sent since 21st May. The cause is shown as "no new messages"but several have been sent to the list in the interim.

The clue seems to be the "* Connection to 'mail.livemail.co.uk' wasn't successful", "* POP3 connection failed!" messages for the Bridge plugin and for the bounce handler.

From the administration control panel, Bridge page, I pressed the on screen button to test the POP3 connection. After a pause, there was a white box containing "This content failed to load". The adjacent manual check button had the same response.

I can log into the POP3 account via the shared host's webmail (using the usual username and password). I found seven email messages sent to the list: three from me on 29th May and 1st and 4th June, three from other subscribers on 10th June and one I sent as a test yesterday. Their presence is consistent with Dada Mail being unable to retrieve them.

Also, in .dada_files/.tmp, there is an mbox text file that seems to have every message since 2019 but finishes with the email received from the non-subscriber on 25th May. I've renamed it to see if it is re-created when the list starts working again.

Wondering why the POP3 access might have stopped, I have been searching through the shared host's support forum. In November 2020, users with Thunderbird were complaining that their client was now using TLS 1.2 by default while livemail.co.uk, the host's mail service, was still on TLS 1. There has since been an undated announcement that livemail.co.uk has changed to TLS 1.2 and that clients unable to use TLS 1.2 may have problems.

Since 2019, I have been using POP3.pm and IO:Socket:SSL v. 2.066. Should I try setting SSL_version to 'SSLv23:!SSLv2:!SSLv3:!TLSv1_1:!TLSv1_2' in POP3.pm, as the metacpan documentation suggests?

Or am I barking up the wrong tree entirely and should I be looking for why the website and possibly the SQL databases are running slowly and causing timeouts, including for POP3 access?

The semaphore messages that I keep logged as a visible heartbeat for the list, mentioned in my November 2019 dadadev emails, changed after 23:40 26th May to "Couldn't create an exclusive semaphore file at ...ScheduledTasks.pm line 229, line 1" without time stamps.

What could have happened near midnight on 26th May, to stop the list retrieving and sending out emails?

Kind regards

Chris