Search results for "sapiro" - lists.mailman3.org

[MM3-users] Re: possible backscatter attack using Mailman3 servers

by Mark Sapiro

On 1/3/25 16:18, David Newman wrote: > Greetings. On a system running Mailman 3.3.9 and Postfix, I'm seeing > about 20-30 entries per day in the Postfix queue where it appears a > Gmail user signs up for a mailing list that requires confirmation, and > Gmail responds that the user is too busy to handle requests. > > There are no publicly advertised email lists on this server, and I don't > ever see anything in the Mailman logs indicating the user ever tried > signing up. This is an attack mail bombing the user. The requests that result in the can come via web or email. Mailman's logging of subscribes has been missing most events through Mailman 3.3.10. See https://gitlab.com/mailman/mailman/-/issues/1143 which will be fixed in 3.3.11, but subscribes waiting user confirmation still won't be logged. However, the message with subject "Please Confirm Your Email Address" comes from Django allauth so it isn't actually Mailman sending it but rather Django allauth as a result of a request to sign up for a Django account at https://mail.example3.com/accounts/signup/. You can probably find that request in your web server logs, and you may find the user and/or email in the Django admin UI. Since django-mailman3 1.3.6, you can disable these signups by putting ACCOUNT_ADAPTER = 'django_mailman3.views.user_adapter.DisableSignupAdapter' in your Django settings, but then your users won't be able to sign up for web accounts. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 5 months

[MM3-users] Re: Uncaught runner exception: 'str' object has no attribute 'decode'

by Mark Sapiro

On 10/02/2017 05:25 AM, Henrik Rasmussen wrote: >> The bottom line is at this point, <https://bugs.python.org/issue27321> >> no longer affects Postorius, but it still affects message delivery. I >> have opened <https://gitlab.com/mailman/mailman/issues/406> about >> defending against that and also the UnicodeEncodeError you included in >> your OP. > > Is there some kind of workaround, I can give my users or something that the list admins can do as a workaround? There is no user or list admin workaround, but the defense is a fairly simple patch. See the MR at <https://gitlab.com/mailman/mailman/merge_requests/322> and the commit at <https://gitlab.com/mailman/mailman/commit/37b3f62afdf225a8e29b66fcefcdb56b9…>. The commit changes 7 files, but only two are needed. The rests are tests or reverting <https://gitlab.com/mailman/mailman/issues/256> which isn't necessary. The only changes you need are those to mailman/email/message.py and mailman/mta/connection.py. >> No. 'bin/mailman queue' is the Mailman 3 command. The Mailman 2 command >> is 'bin/show_qfiles'. > > I much be in the wrong place because when I do a 'docker exec -it mailman-core bash' and do a 'mailman --help' I don't see a 'queue' argument. The only bin/mailman I have inside the mailman-core container is /usr/local/bin/mailman, which is the same command. My bad. The name is 'qfile', not 'queue'. Sorry about that. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

8 years, 8 months

[MM3-users] Re: Archive import fails: "IndexError: list index out of range"

by Mark Sapiro

On 1/7/23 23:04, Stephen J. Turnbull wrote: > Johannes Rohr writes: > > > IndexError: list index out of range > > That isn't the problem, it's the earlier failure, I think: > > > File "/usr/lib/python3.8/email/_header_value_parser.py", line 2069, > > in get_msg_id > > token, value = get_dot_atom_text(value) > > File "/usr/lib/python3.8/email/_header_value_parser.py", line 1334, > > in get_dot_atom_text > > raise errors.HeaderParseError("expected atom at a start of " > > email.errors.HeaderParseError: expected atom at a start of dot-atom-text > > but found '@localhost>' > > It appears that HyperKitty is trying to find a domain name (the common > use for "dot_atom_text" in mail headers in a Message-ID), but is > finding "@localhost>" instead. Actually, this occurs after the message has been converted to an email.message with policy=policy.default. Getting the Message-ID attempts to parse it and in this case, the dot_atom_text it's looking for is the 'local part' which doesn't exist. I.e. <xxx@localhost> would be OK, but <@localhost> is not. > First I would try the check_hk_import script which is provided with > HyperKitty. HyperKitty's contrib/check_hk_import script didn't find this issue. The GitLab branch was updated yesterday, but that update is not yet released. An up to date script which does find this issue is at <https://www.msapiro.net/scripts/check_hk_import>. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 5 months

[MM3-users] Re: Restore/export existing Mailman lists + hyperkitty archives on a new server

by Mark Sapiro

On 7/5/22 12:32 PM, Kimmo L wrote: > Hello! > My plan is to install fresh Mailman3 host, which means that i need to migrate/transfer my existing mailman mailing lists + hyperkitty archives to new host. > Unfortunately i didnt find any good documentation about this. > I tried myself to export and import mailmanweb/mailman databases (from old PostgreSQL to new PostgreSQL) and recreate indexes(mailman-web update_index). Now i can see and search hyperkitty archives, but if try to open one of the mailing list, i get error - HTTP Error 500: {"title": "500 Internal Server Error"} > So im pretty sure, that I'm gonna have to transport some more stuff over, which I have no idea. Maybe existing users and user rights. Dumping the databases from the old serverr and loading them on the new server should work. There should be an exception and traceback logged in Mailman core's mailman.log or possibly in the Django (mailmanweb) log. What is logged? > If its not a big trouble, can someone kindly share, what exactly should i need to export/import and what would be the right steps/commands/moves. I thing just dumping and loading the databases and copying mailman.cfg and the Django settings should do it. Are the domains the same on the new server? If not there may be issues with that. I'd need to see the exception and traceback to say more. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 11 months

[MM3-users] Re: Postfix problem with imported Mailman2 list

by Mark Sapiro

On 3/2/23 13:16, mailman3(a)outofoptions.net wrote: > I installed and tested Mailman3. I imported a Mailman2 list. All looks good but I get this error. > > Recipient address rejected: User unknown in virtual alias table; from=<hit(a)options.org> to=<atm(a)alist.net> > > The aliases seem to show up find in the GUI for the control panel software I use. Postfix virual file has a mapping of atm to atm(a)alist.net . > > It seems mailman at least sees the connection. From the smtp.log under mailman: Presumably there are LHLO, MAIL FROM, and RCPT TO entries before these next ones. > `Mar 02 10:14:13 2023 (1242405) ('127.0.0.1', 56410) recip: atm(a)alist.net > Mar 02 10:14:13 2023 (1242405) ('127.0.0.1', 56410) Data: b'DATA' > Mar 02 10:14:13 2023 (1242405) ('127.0.0.1', 56410) Data: b'QUIT' > Mar 02 10:14:13 2023 (1242405) ('127.0.0.1', 56410) connection lost > Mar 02 10:14:13 2023 (1242405) Connection lost during _handle_client() The connection lost entries are normal in the log. This all looks like Mailman's receiving of the message via LMTP from Postfix. This is inconsistent with Postfix reporting Recipient address rejected: User unknown in virtual alias table which would have occurred before any attempt to relay the message via LMTP. In order to understand more about this, I'd need to see at least the output from `postconf -n`. I'm just guessing here, but https://docs.mailman3.org/projects/mailman/en/latest/src/mailman/docs/mta.h… may be applicable in your case. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

3 years, 3 months

[MM3-users] Re: Hyperkitty Replies showing up as new posts

by Mark Sapiro

On 3/5/25 8:55 PM, Arte Chambers via Mailman-users wrote: > Hello, my Hyperkitty is creating a new post for each reply in a thread. I > would expect that new posts to a list would show up in the archive and once > you click on that thread you would see all replies. However in my > Hyperkitty each reply to a thread shows up as its own post usually with the > subject line saying RE: Original Message. How can I set up Hyperkitty to > behave more like the lists.mailman3.org archives where replies to topics > are filed under their respective threads? HyperKitty's threading is based solely on In-Reply-To: headers. It does not attempt to thread based on References: or Subject: matching. Usually this works, but it can fail if for example A posts to a list. B replies off list to A and A replies on list to B's off list reply. In this case, the In-Reply-To: of A's reply is the Message-ID: of B's reply which isn't on the archive, so A's reply starts a new thread, but this is unusual, and HyperKitty has a "Reattach this thread" feature that a list owner or superuser can use to attach one thread to another. However, if this non-threading occurs regularly, It may be that the replyer's MUA is not setting In-Reply-To:. I've seen some phone clients do this. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year, 3 months

[MM3-users] Re: MM3 + Postfix + Dovecot + PostfixAdmin

by Mark Sapiro

On 8/8/19 10:14 PM, Stephen J. Turnbull wrote: > maztec(a)gmail.com writes: > > > Issue: Postfix LMTP does not pass messages onto MM3:LMTP. > > I don't understand why this is a Mailman problem. Seems like Postfix > user lists would be a better venue to get expert help with this. > Among other things, our most responsive Postfix guy is out of cell > coverage for now. I have a window of availability - and it's longer than I thought because my second of three back-country trips has been canceled :( > I would post this To: the relevant Postfix list, Cc: this list, and > set Reply-To and Mail-Followups-To to the Postfix list to get > everybody who can contribute on the same page. I don't have much to add to what Steve has said except our recommended config includes transport_maps = hash:/path-to-mailman/var/data/postfix_lmtp in Postfix main.cf. This should contain (per your log message) things like ListName-confirm(a)ListServer.TLD lmtp:[127.0.0.1]:8024 to tell Postfix to deliver that address via lmtp to port 8024 on the local host. Dovecot should not be involved. There is something in your config that trumps this and delivers via lmtp to Dovecot for mailman3:[127.0.0.1]:8024. Does your transport_maps look like the above or perhaps like ListName-confirm(a)ListServer.TLD mailman3:[127.0.0.1]:8024 If so, I don't know where mailman3 instead of lmtp: comes from. Also, even if it is lmtp: I don't know that Dovecot can deliver via lmtp. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

6 years, 10 months

[MM3-users] Re: starting gunicorn per mailman.log yet gunicorn not installed

by Mark Sapiro

On 6/26/23 1:15 AM, Odhiambo Washington wrote: > On Mon, Jun 26, 2023 at 9:29 AM Stephen J. Turnbull < > turnbull.stephen.fw(a)u.tsukuba.ac.jp> wrote: > >> Yes, it makes sense. Mailman 3 core is basically an application which >> manages a webserver that provides the REST API, a database of domains, >> users, and lists, an LMTP server for incoming posts and an SMTP client >> for outgoing distribution. Mailman uses gunicorn to provide the >> webserver. >> >> Installation of gunicorn is handled as a package dependency, so you >> probably wouldn't notice unless you were watching the installation >> process very carefully. >> > > It doesn't make sense, IMHO, because as the OP says, > 1. his ExecStart is: > ExecStart=/opt/mailman/venv/bin/uwsgi --ini /etc/mailman3/uwsgi.ini" *and > not* > ExecStart=/opt/mailman/venv/bin/gunicorn -c /etc/mailman3/gunicorn.conf > mailman_web.wsgi:application. > > 2. He has not knowingly created any gunicorn.conf > > Why would Gunicorn be Exec-ing itself? :-) As Steve says, Mailman core uses Gunicorn to manage it's REST API. This has nothing to do with whatever WSGI server (uWSGI, Gunicorn, Apache mod_wsgi, ...) is used to support the Django applications (django-mailman3, hyperkitty and postorius). I.e., there is always a Gunicorn instance configured and started by Mailman core, regardless of whatever WSGI server supports Django or even if one isn't running Django at all. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

2 years, 11 months

[MM3-users] Re: Unable to update templates

by Mark Sapiro

On 5/27/25 18:54, Mark wrote: > > However when I update the template, the databases are updated correctly > but the Welcome message is still sent with the old template content. > > mailman.templates > mailmanweb.postorius_emailtemplate > > Additionally, if I delete the template and create a new one the old > content still gets sent. > > I cannot find where this content is written or is being cached. The content is stored in the mailmanweb.postorius_emailtemplate table. As far as Mailman core and Postorius are concerned, that's the only place. I'm guessing, but my guess is it's cached in your wsgi server. I.e., upon initial creation an entry is made in the mailman.template table with a uri like https://example.com/mailman3/api/templates/list/list.example.com/list:user:… and when Mailman core wants the template, it gets that uri which I'm guessing is then cached in the wsgi server. Then, if the content is changed, the GET returns the old cached value instead of getting the new value from Postorius. If the template is deleted, the mailman.template table entry is deleted and at that point Mailman core will get the template from the file system following it's normal search rules, but if the template is recreated in Postorius, a new table entry is made with the same uri which may still get a cached template from the wsgi server. If this is correct, restarting mailman-web or whatever controls the wsgi server should fix it. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

1 year

[MM3-users] Re: Hyperkitty import from mailman2

by Mark Sapiro

> > Now I wanted to try import from mailman2 mbox, so i did: > > python2 manage.py hyperkitty_import --settings settings --ignore-mtime > -v 3 --since 2010-01-01 -l test(a)list.domain.com /tmp/test.mbox > > It shows that it have sucessfully imported all emails and I can see them > from database (sqlite3), but I cannot see them from GUI (hyperkitty). > Only thing I can see is this inactive "test" list. > > I have tried that I create list before importing, still same result. I have done this several times for several different lists and it has always worked with no problems. I have always created the list first. I do not use any of the options other than -l. My settings location is built in to the django admin script I run. So what I run is equivalent to python2 manage.py hyperkitty_import --settings .. -l .. mbox Do you have cron set up to run the periodic django tasks as in @hourly python2 manage.py runjobs --settings settings hourly @daily python2 manage.py runjobs --settings settings daily @weekly python2 manage.py runjobs --settings settings weekly @monthly python2 manage.py runjobs --settings settings monthly @yearly python2 manage.py runjobs --settings settings yearly 0,15,30,45 * * * * python2 manage.py runjobs --settings settings quarter_hourly * * * * * python2 manage.py runjobs --settings settings minutely If not, did you run python2 manage.py update_index --settings settings after the hyperkitty_import? I'm actually not sure any of this will help since it should only affect the search index, not the message display. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan

8 years, 6 months

Search results for query "sapiro"