Search results for "Postorius" - Mailman-users

Database deadlock when trying to delete a member.

by Simon Coggins

Hi, I’ve got a strange error. I migrated from mailman 2.1 to 3 using the migration guide on the website and things have been working fine. But I’m in the process of testing the new syncmembers command to replace some inhouse mailman API code we wrote. I’ve tracked my problem down to what appears to be a deadlock in the postgresql database. When trying to delete certain users the SQL query hangs and never returns. After doing some tracing and tracking it down to the database, I enabled DB logging of all queries and saw that the following query was never returning: DELETE FROM member WHERE member.id = 166814 If I try to run it manually I get the same issue, in psql it never returns. mailman=# explain analyze DELETE FROM member WHERE member.id = 166814; [… 60 seconds passes…] ^CCancel request sent ERROR: canceling statement due to user request CONTEXT: while deleting tuple (2998,72) in relation "member" I’m just wondering if there are any commands or scripts I can run to verify the database and it’s constraints to see if there is some error in the database data? Or where you’d recommend I go from here? It only impacts the user on this specific list. If I add this user to another list and remove them, everything works fine. This is on a test system so I can do various testing. I’m trying to find the root cause so that I can verify my production system isn’t impacted by the same thing. Setup Details: OS: Oracle Linux 7 (Redhat) Installed using pip via virtualenv talking to a local pgsql database on the same box. Data was migrated from a mailman 2.1 install into mailman3 using migration instructions on the website. PIP Modules versions: django-mailman3 1.3.5 mailman 3.3.3 mailman-hyperkitty 1.1.0 mailmanclient 3.3.2 postorius 1.3.4 psycopg2-binary 2.8.5 RPM packages: postgresql12-12.6-1PGDG.rhel7.x86_64 postgresql12-server-12.6-1PGDG.rhel7.x86_64 postgresql12-libs-12.6-1PGDG.rhel7.x86_64 postgresql12-devel-12.6-1PGDG.rhel7.x86_64 -Simon

4 years, 8 months

Re: Newbie question 2: Rewriting issue for bounced emails

by William Oliver

On Sun, 2021-12-26 at 16:47 -0800, Mark Sapiro wrote: > On 12/26/21 3:06 PM, William Oliver wrote: > > > > When running mailman3, do I need to set up a completely different > > virtual domain for it to run in? > > No. > > > > My "real" domain name is > > libertyfp.org, and I'm trying to make mailman work on that domain, > > e.g. > > "testlist(a)libertyfp.org." I'm starting to get the idea that I need > > to > > stick a completely different domain in there for the mailinglist. > > No. > > DNS or anywhere outside your server. > > Have you actually set an alias domain in Postorius for the > libertyfp.org > domain. It doesn't matter what name it is as long as it isn't > a real domain. A good choice is x.libertyfp.org. Have you set that > or > things like > > testlist(a)libertyfp.org testlist(a)x.libertyfp.org > testlist-bounces(a)libertyfp.org testlist-bounces(a)x.libertyfp.org > testlist-confirm(a)libertyfp.org testlist-confirm(a)x.libertyfp.org > ... [snip] Ah. I misunderstood you. I thought you meant that I had to tweak *postfix* to get this done. It's not that you weren't clear, it's that when I read the mailman3 instructions, I got the impression that *mailman* would set the alias domain automatically, so I was dead set on munging postfix. This makes more sense, and I can back out of some of the changes I made in postfix, which for the last couple of rounds was mostly flailing about randomly -- I've heard that's often not a good admin strategy. So.. if I need to do it myself, can you point me to what file I need to create/modify in the /opt/mailman/mm tree to set up the alias? Thanks! billo

3 years, 11 months

Re: spanish localization

by Victoriano Giralt

El sáb, 13-02-2021 a las 16:44 -0800, Mark Sapiro escribió: > On 2/13/21 4:08 PM, Victoriano Giralt wrote: > > Mark, please, how do you check when translations happened? Nov 7th is a > > very significant date for me (it is my wedding anniversary) and I'm 90% > > sure that I dedicated time to Weblate after that date... > > With respect to Mailman core, all I'm doing is downloading the .po for > Spanish and comparing it to mailman/messages/es/LC_MESSAGES/mailman.po > from the Mailman core 3.3.3 package on PyPI and they are identical. > > If there are translated strings on weblate that are different from that, > they don't show up in the downloaded > <https://hosted.weblate.org/download/gnu-mailman/mailman/es/>;. And you are very right and I have fallen as a newbee ... and first of all I apologise for the noise. I've been doing diffs with files downloaded from Weblate and those installed on the server by pip. As I should have done from the start. Mailman core and hyperkitty show no differences. Postorius shows a bunch and django_mailman3 just a couple. I leave for others a more detailed check of the strings between downloaded .po files and Weblate online strings, it's too late here. > > And I too want to help. It would be a help to me if you can give me > examples of specific strings whose weblate translation differs from > what's in the distributed .po files. And you have helped me, making me thing twice and stop moaning and going for the facts. Apologetic-thankful-big-mouth-shut-ly, Victoriano. -- Victoriano Giralt Innovation Director Digital Transformation Vicerectorate University of Malaga +34952131415 SPAIN ================================================================== Note: signature.asc is the electronic signature of present message A: Yes. > Q: Are you sure ? >> A: Because it reverses the logical flow of conversation. >>> Q: Why is top posting annoying in email ?

4 years, 9 months

Re: Localization of auto-responses, question marks in welcome message

by Stephen J. Turnbull

Claus-Justus Heine via Mailman-users writes: > Kind thanks for the response. I have to apologize: my "test user" had > its preferred language set to en_US, so that was the reason for the > question marks. So this was indeed a user-error (i.e. my fault). But we'd like to have that user get the English version of the message. One of two things seems reasonable: 1. You have customized the German template, but not the English template. In that case, I think it makes sense for Mailman to assume the German template has important information that the English does not, and send the German template with its appropriate encoding (UTF-8, nowadays). The fact that en_US can be encoded in ASCII should not affect the encoding of the de template, ever.[1] 2. You have not customized the German template. In that case the user should get the default English template. How do those sound to you, as goals for Mailman behavoir? So ISTM this is a Mailman issue, if Mailman is sending out the German message with question marks substituted for non-ASCII characters, or if Mailman is sending out the German message with US-ASCII as the Content-Type charset parameter. The only way I can see this as a user (admin) issue is if you overwrote the English template with German (and even then, it's nuts for Mailman not to use UTF-8 for English what with directed quotes, emoji, and all the other enhancements to the character repertoire that US English speakers use). Can you tell me a little more about whether you configured any custom templates, and if so, how? That is, did you use the Postorius text field or overwrite files directly? Steve Footnotes: [1] Note to self: perhaps the admin UI can check the list of supported languages and warn if some but not all are customized.

3 years, 7 months

Race condition when mass subscribing same email addresses to two lists?

by Alan So

We are testing the latest Mailman core 3.3.5 and Postorius 1.3.6. My colleagues reported that some email addresses cannot subscribe to the lists when two of them are using mass subscribe feature to add email addresses to the list. Then I try whether I can repeat the same issue. I did this: 1) Create two lists with default setting 2) Prepare a list of 1000 fresh email addresses not added to the system before 3) Mass subscribe both lists of these 1000 email addresses around the same time with all three Pre (confirm, approved, verified) checked There are quite a lot of similar errors in mailman.log. Did not study the code yet. Which part is causing the issue? SQLAlchemy 1.3.24 psycopg2-binary 2.8.6 [2021-10-28 08:38:30 +0000] [20141] [ERROR] Error handling request /3.1/members Traceback (most recent call last): File "/opt/mailman/venv/lib/python3.6/site-packages/gunicorn/workers/sync.py", line 136, in handle self.handle_request(listener, req, client, addr) File "/opt/mailman/venv/lib/python3.6/site-packages/gunicorn/workers/sync.py", line 179, in handle_request respiter = self.wsgi(environ, resp.start_response) File "/opt/mailman/venv/lib/python3.6/site-packages/mailman/database/transaction.py", line 51, in wrapper config.db.commit() File "/opt/mailman/venv/lib/python3.6/site-packages/mailman/database/base.py", line 52, in commit self.store.commit() File "/opt/mailman/venv/lib/python3.6/site-packages/sqlalchemy/orm/session.py", line 1046, in commit self.transaction.commit() File "/opt/mailman/venv/lib/python3.6/site-packages/sqlalchemy/orm/session.py", line 502, in commit self._assert_active(prepared_ok=True) File "/opt/mailman/venv/lib/python3.6/site-packages/sqlalchemy/orm/session.py", line 296, in _assert_active code="7s2a", sqlalchemy.exc.InvalidRequestError: This Session's transaction has been rolled back due to a previous exception during flush. To begin a new transaction with this Session, first issue Session.rollback(). Original exception was: (raised as a result of Query-invoked autoflush; consider using a session.no_autoflush block if this flush is occurring prematurely) (psycopg2.errors.UniqueViolation) duplicate key value violates unique constraint "ix_address_email" DETAIL: Key (email)=(testemail999(a)example.com) already exists.

4 years, 1 month

Initial success: Mailman 3.3 integrated with Zimbra 8.8

by Morris Jones

I have a very old existing mail server running Zimbra 6 with a mailman 2.1 instance alongside. I needed to rebuild my mail server for some compelling reasons (the old server used outdated TLS protocol, I could no longer compile clamav, nothing could really be updated). The final straw was a lack of support for my old commercial SSL certificates. I patched in a letsencrypt certificate to get through the past six months. I'm in the process of commissioning a new mail server using an 8GB Linode, Ubuntu 18, Zimbra 8.8, and Mailman 3.3. A couple of days ago I had my first successful list creation, with user confirmation working by email. I have django running with uwsgi, and am using Zimbra's installed nginx proxy to reverse proxy http traffic to Postorius and Hyperkitty. Nginx is doing the SSL layer. The mailman core mta hooks are all done using Zimbra's "local configuration" technique. In theory I should be able to do Zimbra upgrades with a minimum of hassle. I've written up a "case history" of how it came together. I have a .rst file in a gitlab project here: https://gitlab.com/MojoLA/mailman3.3-zimbra-8.8-integration I suppose this is a crazy project, but it's worthwhile to me. (Mark, thank's so much for integrating the virtual domain maps. It took me about three days to recognize that they were the solution to bounced mails to list-owner@...) Now I can spend some time migrating my email accounts (only four) from the old server to the new one, and move a dozen or so mailing lists to the new mailman3 instance. I'll have to get accustomed to a new way to manage the mail lists, and already really like what I've seen. Thank you. :) Best regards, Mojo -- Morris Jones, Monrovia, CA BridgeMojo <http://bridgemojo.com> Old Town Sidewalk Astronomers <http://otastro.org> Mojo's Blog <http://mojo.whiteoaks.com>

5 years

Re: Strange password problem

by Eggert Ehmke

There may be broken entries in my config. In web/logs/uwsgi-error.log I get: Unknown command: 'qcluster' when the entry in uwsgi.ini is: *attach-daemon* *=* python3 /opt/mailman/venv/lib/python3.9/site-packages/ mailman_web/manage.py qcluster When I change that line to: *attach-daemon* *=* python3 ./manage.py qcluster I get: python3: can't open file '//./manage.py': [Errno 2] No such file or directory This may be unrelated to the password problem, but this line runs every minute in the cron job. Am Donnerstag, 22. Juni 2023, 11:50:57 CEST schrieb Odhiambo Washington: > On Thu, Jun 22, 2023 at 12:34 PM Eggert Ehmke via Mailman-users < > mailman-users(a)mailman3.org> wrote: > > > > This happened several times recently with Postorius Version 1.3.8: > > -I try to log in as admin > > -rejected because of wrong password (password is stored in browser) > > -used forgotten password function > > -reset password to the stored value from browser > > -login works now. Works also when logged out/in again > > -however, the next day this behaviour repeats. > > > > > > > > Any hints? > > > > 1. You have a bug in the browser cache. Try a different browser. > 2. You can check also mailmanweb logs for clues?? > > -- > Best regards, > Odhiambo WASHINGTON, > Nairobi,KE > +254 7 3200 0004/+254 7 2274 3223 > "Oh, the cruft.", egrep -v '^$|^.*#' ¯\_(ツ)_/¯ :-) > [How to ask smart questions: > http://www.catb.org/~esr/faqs/smart-questions.html] > _______________________________________________ > Mailman-users mailing list -- mailman-users(a)mailman3.org > To unsubscribe send an email to mailman-users-leave(a)mailman3.org > https://lists.mailman3.org/mailman3/lists/mailman-users.mailman3.org/ > Archived at: > https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message > /BWAVUEMHI56YI7FHTDULIZNIKIAM7324/ > > This message sent to eggert(a)eehmke.de ___________________________________________ Mailman's content filtering has removed the following MIME parts from this message. Replaced multipart/alternative part with first alternative.

2 years, 5 months

Re: Confirmation emails to Users has wrong domain name (example.com!)

by Abhilash Raj

On Thu, Sep 30, 2021, at 2:20 AM, Odhiambo Washington wrote: > On Thu, Sep 30, 2021 at 1:59 AM Abhilash Raj <maxking(a)asynchronous.in> > wrote: > >> >> >> > On Sep 29, 2021, at 2:34 PM, Odhiambo Washington <odhiambo(a)gmail.com> >> wrote: >> > >> > On Wed, Sep 29, 2021 at 8:31 PM Mark Sapiro <mark(a)msapiro.net> wrote: >> > >> >> On 9/29/21 9:50 AM, Odhiambo Washington wrote: >> >>> 1. Confirmation emails to Users has wrong domain name (example.com!) >> >>> <https://docs.mailman3.org/en/latest/faq.html#id1> >> >>> < >> >> >> https://docs.mailman3.org/en/latest/faq.html#confirmation-emails-to-users-h… >> >>> >> >>> >> >>> This happens when your reverse (SSL) proxy isn’t setting up the correct >> >>> headers when proxying requests. Fix this by setting the right >> >>> proxy_set_header directives: >> >> ... >> >>> How is this supposed to be mitigated in Apache when using WSGI? >> >>> >> >>> My config: >> >>> >> >>> WSGIDaemonProcess hyperkitty threads=25 python-path=/usr/local/mailman >> >>> user=mailman group=mailman >> >>> WSGIPythonHome "/usr/local" >> >>> WSGIProcessGroup hyperkitty >> >> >> >> >> >> You are using mod_wsgi and not proxying at all, so this is not relevant >> >> in your case. >> >> >> >> Are you actually seeing this issue? If so, it might be related to >> >> >> >> >> https://docs.mailman3.org/en/latest/faq.html#the-domain-name-displayed-in-h… >> >> >> > >> > My issue is related to this, but the documentation referred to is not for >> > the faint-hearted! >> > I can't make head or tails of it. >> >> Click on “Domain” in Postorius from the top bar, which should take you to >> the Domains page. >> >> For your domain (in the “Mail Host” column), see the corresponding “Web >> Host” column, it should look something like: >> >> lists.mailman3.org (lists.mailman3.org) >> (Edit) >> SITE_ID = 1 >> > > Mine looks different, slightly. But there is no example.com at all. > > [image: Abhilash.png] > > > >> >> on a new-ish version of Postorius. >> > > I have the newest versions of everything, having installed only yesterday. > > >> >> If it doesn’t show the right values and instead shows “example.com” for >> you, click on the “Edit” link, which will take >> you to a page that will allow you to edit both the values. >> > > It shows the right values, but with "SITE_ID = 2". In my settings_local.py > I have SITE_ID = 1. > I suppose the example.com is the one tied to SITE_ID = 1 and that is what I > have in my settings_local.py. That is most likely what is happening. > Should I edit my settings_local.py? Yes, please update it to have SITE_ID = 2 to correspond to the site you want and restart. > > I am still confused. > > -- > Best regards, > Odhiambo WASHINGTON, > Nairobi,KE > +254 7 3200 0004/+254 7 2274 3223 > "Oh, the cruft.", egrep -v '^$|^.*#' :-) > _______________________________________________ > Mailman-users mailing list -- mailman-users(a)mailman3.org > To unsubscribe send an email to mailman-users-leave(a)mailman3.org > https://lists.mailman3.org/mailman3/lists/mailman-users.mailman3.org/ -- thanks, Abhilash Raj (maxking)

4 years, 2 months

Re: nginx configuration on a multitasking server

by David Newman

On 12/22/21 8:50 PM, Mark Sapiro wrote: > On 12/22/21 7:54 PM, David Newman wrote: >> On 12/22/21 7:29 PM, Mark Sapiro wrote: >> >>> This looks like a uwsgi issue as discussed at >>> https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/thread/… >>> although that's with apache and not nginx so it may not be. >> >> If I'm reading the thread correctly, it's Apache workarounds to an >> underlying Debian issue, so I agree it may not apply with nginx. > > > I think you're correct that it doesn't apply to nginx, but the symptom > is not the same. It is not a Debian issue per se. It is an Apache change > that got installed with a Debian update, but the underlying issue is > with Apache. See > https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/message… > > > >> There are 260 files named 'urls.py' under /opt/mailman, mostly for >> socialaccount stuff. >> >> Where does that change go? All files? Or just this one: >> >> /opt/mailman/venv/lib/python3.9/site-packages/postorius/urls.py > > > Not that one. The one whose content looks like > https://gitlab.com/mailman/mailman-web/-/blob/master/mailman_web/urls.py. It > is probably in the same directory as your Django settings. On my system that file lives here: /opt/mailman/venv/lib/python3.9/site-packages/mailman_web/urls.py But the error persists after this addition and restarting MM3 and mailmanweb and nginx: url(r'^mailman/', include('postorius.urls')), There is this 404 error in uwsgi.log: [pid: 73356|app: 0|req: 3/3] 127.0.0.1 () {48 vars in 1544 bytes} [Wed Dec 22 21:13:48 2021] GET /mailman/admindb/listname => generated 5328 bytes in 14 msecs (HTTP/1.0 404) 6 headers in 197 bytes (1 switches on core 0) If this is a uwsgi issue, is there something else to tweak with that? Thanks dn > >

3 years, 11 months

Re: mailman addmember/syncmember commands print undecoded UTF-8 [ADD] or [DEL] mail addresses in its messages like "[ADD] =?utf-8?q?John_=....."

by Wikinaut

Not a single mail was sent - I am migrating (and newly setting up from textfiles) the mailman3 suite, installed fresh from the scratch in (venv) in /opt/mailman. I noticed that after having used a mix of addmember and also syncmember, one (or yesterday about 100 entries) of 500 had the MIME-coding. Then I used a script to remove users with the "=?UTF...." begin. In postorius users the user was gone and also in the lists where this user was in. def delete_users(): url = f"{API_URL}/users" r = requests.get(f'{url}', auth = AUTH) r.raise_for_status() users = r.json()['entries'] for user in users: dn = user.get('display_name') if type(dn) == type("string") and dn.startswith("="): print (dn) self_link = user.get('self_link') requests.delete(self_link, auth=AUTH) def main(): delete_users() After a new add/syncmember, that spooky entry was up again! Then I used mailman delmembers --fromall --no-goodbye-msg --no-admin-notify -m <this@address>: Then after a new add/syncmember, that entry was gone and the correct entry was shown. That means, that somewhere some remains were stored somewhere. I also have a another issue, where I cannot get rid of a test display_name like "1111 [äöo]" <myemailaddress> for my own user entry (this issue has not yet been filed, as I do not want to flood you here; the "=?utf.." issue seems to be a more apparent problem). At the moment, none of the ~500 users have "utf" searchable, which is some good news. Perhaps to mention: I noticed the utf-problems when I used Ctrl-C during heavy script actions while add/syncing the lists which I needed to stop. while testing. Perhaps this caused an inconsistent database state??? and that present issue was a consequence of my harsh break?

3 months, 4 weeks

Mailman-users search results for query "Postorius"