hi,

today i finally migrated my mailman3 instance from Debian/bookworm to Debian/trixie, using the Debian packages of mailman3. everything worked great, and i'm now running:

Mailman Core Version: GNU Mailman 3.3.10 (Tom Sawyer) Mailman Core API-Version: 3.1 Mailman Core Python-Version: 3.13.5 (main, Aug 10 2026, 12:06:59) [GCC 14.2.0]

the Debian packages have the instance is backed by a local PostgreSQL-17 instance (17.11.0)

• python3-django-hyperkitty 1.3.12
• python3-django-postorius 1.3.13
• python3-psycopg2 2.9.10

during the upgrade, I noticed that postgresql is constantly writing warnings to its logfile, like so:

2026-10-29 17:26:06.875 CET [171290] [unknown]@[unknown] LOG: could not accept SSL connection: EOF detected

"constantly" meaning: about 4.5 times per second! or ~16400 lines per hour. this is obviously triggered by the webinterface, as the errors stop immediately if i issue a systemctl stop mailman3-web.

it's also notable, that I do not notice any actual problems on the webinterface, including the archives (so the DB connection seems to work fine). accordingly, /var/log/mailman3/web/mailman-web.log doesn't indicate any problems.

also, this issue is not something introduced by the upgrade: i got the same warnings with my old mailman3.3.8 instance with PostgreSQL-15. it's just that i noticed the errors only now.

here's the relevant section (afaict) from /etc/mailman3/mailman-web.py:

DATABASES = {
     'default': {
         'ENGINE': 'django.db.backends.postgresql_psycopg2',
         'NAME': 'mailman3web.db',
         'USER': 'REDACTED',
         'PASSWORD': 'REDACTED',
         'HOST': 'localhost',
         'PORT': '',
         'OPTIONS': {
         },
     }
}

the ssl related settings in my /etc/postgresql/17/main/postgresql.conf are (these are just the defaults, i didn't change anything):

ssl = on
ssl_cert_file = '/etc/ssl/certs/ssl-cert-snakeoil.pem'
ssl_key_file = '/etc/ssl/private/ssl-cert-snakeoil.key'

since everything appears to work, i might ignore the issue, but since the logfiles get pretty big due to the permanent logging, i'd rather not...

any ideas what is going on?

-- IOhannes m zmölnig IEM network operation center https://noc.iem.at/