- Mailman-users - lists.mailman3.org

Mailman 3.3.4 logging behaviour, mixed log formats, and REST API error
by Henrik Rasmussen Aug. 4, 2026

Aug. 4, 2026

Hello, I have been chatting with Copilot AI for some days to get help investigating some logging-related issues in a Mailman 3.3.4 Docker installation and, so I would appreciate any guidance. We are working on a long needed upgrade, but for now I need the logs in Elasticsearch, and they are aweful. Environment * GNU Mailman 3.3.4 (Tom Sawyer) * Python 3.8.10 * Docker-based installation * Test environment with a single mailing list: listname(a)example.com<mailto:listname@example.com> Logging configuration I've configured dedicated log files for most logging channels. For example, [logging.http] is configured with format: %(asctime)s (%(process)d) %(message)s, datefmt: %Y-%m-%d %H:%M:%S.%3N, propagate: no, level: debug, and path: http.log. All [logging.*] sections in the generated /etc/mailman-extra.cfg use the same datefmt. format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: info path: mailman.log [logging.archiver] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: archiver.log [logging.bounce] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: bounce.log [logging.config] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: config.log [logging.database] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: database.log [logging.debug] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: debug.log [logging.error] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: error.log [logging.fromusenet] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: fromusenet.log [logging.http] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: http.log [logging.locks] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: locks.log [logging.mischief] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: mischief.log [logging.plugins] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: plugins.log [logging.runner] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: info path: runner.log [logging.smtp] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: info path: smtp.log [logging.smtp-failure] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: smtp-failure.log [logging.subscribe] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: subscribe.log [logging.vette] format: %(asctime)s (%(process)d) %(message)s datefmt: %Y-%m-%d %H:%M:%S.%3N propagate: no level: debug path: vette.log Issue 1: Mixed log formats Despite configuring all logging sections with datefmt: %Y-%m-%d %H:%M:%S.%3N, I still see multiple timestamp formats in mailman.log, for example: * [30/Jul/2026:12:00:03 +0000] "GET /3.1/lists?count=10&page=1 HTTP/1.1" 200 90 "-" "GNU Mailman REST client v3.3.3" * [2026-07-30 12:32:15 +0000] [36] [INFO] Shutting down: Master * Jul 30 12:32:14 2026 (1) Master watcher caught SIGTERM. Exiting This suggests that multiple logging systems may be writing to the same log file. Issue 2: Runner debug noise With logging level set to debug, mailman.log contains a large number of entries such as: * Jul 31 07:58:12 2026 (33) [PipelineRunner] starting oneloop * Jul 31 07:58:12 2026 (33) [PipelineRunner] ending oneloop: 0 * Jul 31 07:58:12 2026 (26) [ArchiveRunner] starting oneloop * Jul 31 07:58:12 2026 (26) [ArchiveRunner] ending oneloop: 0 * Jul 31 07:58:12 2026 (29) [IncomingRunner] starting oneloop * Jul 31 07:58:12 2026 (29) [IncomingRunner] ending oneloop: 0 * Jul 31 07:58:13 2026 (28) [CommandRunner] starting oneloop * Jul 31 07:58:13 2026 (36) [VirginRunner] starting oneloop * Jul 31 07:58:13 2026 (36) [VirginRunner] ending oneloop: 0 * Jul 31 07:58:13 2026 (28) [CommandRunner] ending oneloop: 0 * Jul 31 07:58:13 2026 (37) [DigestRunner] starting oneloop * Jul 31 07:58:13 2026 (37) [DigestRunner] ending oneloop: 0 * Jul 31 07:58:13 2026 (31) [NNTPRunner] starting oneloop * Jul 31 07:58:13 2026 (32) [OutgoingRunner] starting oneloop * Jul 31 07:58:13 2026 (31) [NNTPRunner] ending oneloop: 0 * Jul 31 07:58:13 2026 (32) [OutgoingRunner] ending oneloop: 0 Is there a recommended way to suppress these internal runner loop messages while retaining useful operational logging? Issue 3: HTTP request logging I intentionally keep [logging.http] at level: debug because I want to see which REST endpoints are called. Examples: * [2026-08-03 12:10:28 +0000] [51] [DEBUG] GET /3.1/lists/listname(a)example.com<mailto:/3.1/lists/listname@example.com> * [2026-08-03 12:15:29 +0000] [52] [DEBUG] GET /3.1/lists/listname(a)example.com<mailto:/3.1/lists/listname@example.com> * [2026-08-03 13:00:03 +0000] [53] [DEBUG] GET /3.1/lists My goal is to keep these request entries. Issue 4: Falcon tracebacks in http.log The same http.log also contains Falcon tracebacks. One example is: 2026-08-03 12:10:29 [FALCON] [ERROR] GET /3.1/lists/listname(a)example.com/config<mailto:/3.1/lists/listname@example.com/config> => Traceback (most recent call last): File "/usr/lib/python3.8/site-packages/sqlalchemy/engine/result.py", line 779, in _getter getter = self._metadata._getter AttributeError: 'NoneType' object has no attribute '_getter' The above exception was the direct cause of the following exception: Traceback (most recent call last): File "/usr/lib/python3.8/site-packages/falcon/app.py", line 361, in call responder(req, resp, **params) File "/usr/lib/python3.8/site-packages/mailman/rest/listconf.py", line 282, in on_get value = getter.get(self._mlist, attribute) File "/usr/lib/python3.8/site-packages/mailman/rest/listconf.py", line 52, in get return sorted(aliases.aliases) File "/usr/lib/python3.8/site-packages/mailman/model/mailinglist.py", line 565, in aliases for alias in aliases File "/usr/lib/python3.8/site-packages/sqlalchemy/orm/loading.py", line 100, in instances cursor.close() File "/usr/lib/python3.8/site-packages/sqlalchemy/util/langhelpers.py", line 68, in exit compat.raise_( File "/usr/lib/python3.8/site-packages/sqlalchemy/util/compat.py", line 182, in raise_ raise exception File "/usr/lib/python3.8/site-packages/sqlalchemy/orm/loading.py", line 58, in instances *[ File "/usr/lib/python3.8/site-packages/sqlalchemy/orm/loading.py", line 59, in query_entity.row_processor(query, context, cursor) File "/usr/lib/python3.8/site-packages/sqlalchemy/orm/query.py", line 4422, in row_processor _instance = loading._instance_processor( File "/usr/lib/python3.8/site-packages/sqlalchemy/orm/loading.py", line 421, in _instance_processor getter = result._getter(col, False) File "/usr/lib/python3.8/site-packages/sqlalchemy/engine/result.py", line 781, in _getter return self._non_result(None, err) File "/usr/lib/python3.8/site-packages/sqlalchemy/engine/result.py", line 1236, in non_result util.raise( File "/usr/lib/python3.8/site-packages/sqlalchemy/util/compat.py", line 182, in raise_ raise exception sqlalchemy.exc.ResourceClosedError: This result object does not return rows. It has been closed automatically. Is it possible to: 1. Continue logging HTTP GET requests to http.log 2. Send Falcon tracebacks and REST API errors to a separate log file At the moment, both request logging and tracebacks appear in the same file. Issue 5: Worker timeout After the traceback above, I consistently see: * [2026-08-03 12:21:29 +0000] [34] [CRITICAL] WORKER TIMEOUT (pid:52) * [2026-08-03 12:21:30 +0000] [53] [INFO] Booting worker with pid: 53 This appears to indicate that the worker handling GET /3.1/lists/listname(a)example.com/config<mailto:/3.1/lists/listname@example.com/config> times out and is restarted. Additional observations error.log contains Mailman Core exceptions such as: (52) deque: subscribe_from_restored Traceback (most recent call last): File "/usr/lib/python3.8/site-packages/mailman/app/workflow.py", line 69, in next return step() File "/usr/lib/python3.8/site-packages/mailman/app/subscriptions.py", line 341, in _step_get_moderator_approval raise StopIteration StopIteration However, the Falcon/REST traceback shown above does not appear in error.log. This suggests that: * error.log receives Mailman Core exceptions. * http.log receives REST API/Falcon exceptions. * The two do not appear to be linked. Questions 1. Is the mixed timestamp formatting expected in Mailman 3.3.4 Docker deployments? 2. Which component produces the starting oneloop / ending oneloop runner messages? 3. Can HTTP request logging and Falcon tracebacks be directed to separate log files? 4. Is ResourceClosedError when accessing GET /3.1/lists/listname(a)example.com/config<mailto:/3.1/lists/listname@example.com/config> a known Mailman 3.3.4 issue? 5. Is there a recommended logging configuration that preserves REST API access logging while reducing runner/debug noise? Thanks in advance. Best regards, Henrik

1 0

Advice on upgrading..
by Stephen Daniel Aug. 4, 2026

Aug. 4, 2026

Mailman 3 people -- I am planning an upgrade for my Mailman 3 system, and I have a couple of questions. I am currently running an old install of Mailman 3.3.5 on Debian Bullseye. Rather than trying to upgrade this now ancient VM, my plan is to create a new VM and perform a fresh install of Mailman 3.3.10 on Debian Trixie. I'd like advice on how to migrate mailing lists and archives from the old install to the new one. Is there a standard procedure? My current implementation uses the Python venv installation method. I'd prefer a more container-focused installation, but since I don't believe Mailman 3 has official container images yet, I plan to continue with the venv approach. Please let me know your thoughts. Thanks! -- Stephen Daniel

4 3

Mass change Digest delivery mode
by iMark Aug. 4, 2026

Aug. 4, 2026

Hi All, Using the Mailman shell, can the delivery mode of Digests be changed to 'Plain text' for all users on just one list ? If it is, could someone tell me how that's done. Thanks, iMark

3 4

Find a smtp server to send out emails
by Leon Aug. 1, 2026

Aug. 1, 2026

Hi, This problem is not caused by mailman, but I still want to give it a shot here. I'm hosting a mailing list on top of mailman3. Emails are supposed to be sent out by AWS ses. However, ses requires that sender address must be verified, which leads to a problem that emails sent by subscribers to mailing list cannot be sent to other subcirbers, since their addresses are not verified. And it is impossible to verified every subscriber. Are there smtp service providers allowing unverified email address to send out emails, or do I have other solutions? Thanks. -- Best wishes, Leon

4 4

FutureWarning: Use `header_id_prefix` instead
by Andreas Vetter July 28, 2026

July 28, 2026

Hi, after a recent update of the prerequisites using: pip install -U --upgrade-strategy eager mailman postorius django-mailman3 hyperkitty mailman-web I get warnings from the cronjobs: /opt/mailman/venv/lib64/python3.11/site-packages/readme_renderer/ markdown.py:44: FutureWarning: Use `header_id_prefix` instead gfm_extension_options.header_ids = _HEADER_ID_PREFIX What would be the best way to silence the warnings? Google does not find the warning message yet. Cheers, Andreas

1 2

Disable signup-requests for mailman3 via e-mail
by Stefan Bauer July 24, 2026

July 24, 2026

Hi Mailman-Experts, how can i completely disable/block list-signup-requests for specific lists via e-mail? i have some announce-only mailinglists i do not want to allow users to signup at all via e-mail. I'm aware of the subscription policies but there is no block or reject/discard option. Thank you. Stefan

3 2

Shutdown problem
by Lars Bjørndal July 23, 2026

July 23, 2026

Sometimes when I reboot the machine, I experience that mailman doesn't start because the files /opt/mailman3/mm/var/locks/master.lck* isn't deleted during shutdown. Maybe one way to solve this could be to use ExecStartPre directive in the systemd service to remove the lock files. What's the best practice to solve this problem? BTW I've noticed that doing `systemctl stop mailman3` takes a very long time, 45 secs. or something like that. - Mailman 3.3.10 - Python 3.12.13 ``` /etc/systemd/system/mailman3.service [Unit] Description=GNU Mailing List Manager After=syslog.target network.target postgresql.service [Service] Type=forking PIDFile=/opt/mailman3/mm/var/master.pid User=mailman3 Group=mailman3 Environment="MAILMAN_CONFIG_FILE=/etc/mailman3/mailman.cfg" ExecStart=/opt/mailman3/venv/bin/mailman start ExecReload=/opt/mailman3/venv/bin/mailman restart ExecStop=/opt/mailman3/venv/bin/mailman stop # Tested, but have no effect on deletion of the lock file. #Restart=on-failure #RestartSec=15 [Install] WantedBy=multi-user.target ``` Lars

6 17

Using Hyperkitty to Create Archive from mbox file
by jason＠glazer350.com July 16, 2026

July 16, 2026

My web host provider (Dreamhost) is shutting down support for Mailman 2. I am planning on transitioning some of the active mailing lists, but for the older ones, I was hoping to create archives of them. I am considering just making a mirror of the "pipermail" archives, using MHonarc, or trying to use Hyperkitty. I am most impressed by the layout and usability of Hyperkitty archives and would like to go that route if it is not too difficult. What would I need to do to take the mbox archive from a mailing list and convert it into an archive using Hyperkitty? Jason

3 3

Find all mailing lists a user is subscribed to
by Rodolfo Pilas July 10, 2026

July 10, 2026

Hi, I administer a Mailman 3 server with more than 50 mailing lists. All subscribers belong to the same organization. Occasionally I need to answer questions such as: “Which mailing lists is user(a)example.com subscribed to?” At the moment I iterate over every mailing list, list its members, and grep for the email address. It works, but it is inefficient. Is there a supported way (CLI, REST API, or Postorius) to retrieve all the mailing lists a particular email address is subscribed to, without having to inspect every list individually? Rodolfo Pilas

2 1

Odd behavior: notifications for list membership, etc. from "owner user" sent as "owner user"?
by Thomas Ward July 10, 2026

July 10, 2026

Tracing down an issue of "why are we no longer receiving owner notices" such as list membership changes, moderation list items, etc. we are seeing errors in MS365 side of things about "failing to deliver". Upon deeper trackdown of this, our owner user mailman(a)example.com is seeing messages sent to it from @lists.example.com and @distro.example.com from mailman(a)example.com (despite the From in the envelope being list-bounces+mailman=example.com(a)[lists|distro].example.com) Is there any specific reason for this? Is it because the MS365 side is ignoring the Envelope's From address and relying on the actual *from* field which is breaking things because MS365 users are "not allowed to send as [lists|distro].example.com"? This seems highly suspicious if this is the case, and I want to see if there's anything we can do on the Mailman side of things to make this 'work' for the 365 recipient (mailman(a)example.com is a shared mailbox for the IT team in this case, so it's a legitimate user).

3 6