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

[MM3-users] Re: UnpicklingError on mailman lists command

by Bryan Loniewski

Okay, so the issue has been fixed by updating the data to an empty list equivalent (after finding, confirming and copying another lists empty value): MySQL [mailman]> SELECT accept_these_nonmembers INTO @newatn FROM mailinglist WHERE list_name='foo_is_a_good_list_with_emptyval'; MySQL [mailman]> UPDATE mailinglist SET accept_these_nonmembers=@newatn WHERE list_name='the_borked_list'; Query OK, 1 row affected (0.02 sec) Rows matched: 1 Changed: 1 Warnings: 0 Thanks for the help on troubleshooting that! Based on your import21 comments, I'm realizing the issue may be caused by multiple runs of import21. I have been (although I've stopped it for some time now) running import21 nightly to sync up my Mailman3 list data with Mailman2. I've been doing the same for archiving as well. I've noticed now that some lists have duplicate acceptable alias entries (looks to be like <number of aliases> x <number of import21 runs>) and the same goes for *_these_nonmembers entries. Is this a symptom of running the old(er) import21 command pre-3.3.2? Will the 3.3.2 import21 work as expected w/out the duplications if I was to clear out the appropriate lists col data (i.e., is it okay to run import21 multiple times)? I tried to figure out if there were any other duplicative data, but did not come across any. Is there more that I'm missing? ________________________ Bryan Loniewski Rutgers University OIT:EAS:Messaging and Collaboration Services System Administrator ________________________________________ From: Mark Sapiro <mark(a)msapiro.net> Sent: Tuesday, August 3, 2021 12:28 PM To: mailman-users(a)mailman3.org Subject: [MM3-users] Re: UnpicklingError on mailman lists command On 8/3/21 7:35 AM, Bryan Loniewski wrote: > Okay, that helped identify possible culprits. Three of our lists are hitting/over the max BLOB length (65,535 bytes) for accept_these_nonmembers. I executed this SQL command for each of the BLOB cols you referenced and only accept_these_nonmembers is borked: > > MySQL [mailman]> SELECT list_name, OCTET_LENGTH(accept_these_nonmembers) as howManyBytes from mailinglist where OCTET_LENGTH(accept_these_nonmembers) > 64000; > +---------------+--------------+ > | list_name | howManyBytes | > +---------------+--------------+ > | <snip> | 65535 | > | <snip> | 65535 | > | <snip> | 65535 | > +---------------+--------------+ > 3 rows in set (0.01 sec) > > So, putting aside *why* these lists are so large wrt that column, can or should I alter the col type(s) to MEDIUMBLOB? If yes, would that remain on suite upgrades? If no, is there a better alternative? Or some recommendation in between? Modifying the column type at this point will not help because the data has already been truncated and the missing piece is gone. It may help avoid future issues, but I suspect it's not necessary. It would probably 'stick' on an upgrade as long as there were no new migration affecting those columns, but if there were such a migration, it could fail due to the existing type not being as expected. I suspect these lists were imported from Mailman 2.1 with the `mailman import21` command. The goal of that is to import the addresses as non-members with the appropriate action and put only the regexps in the MM 3 *_these_nonmembers. Prior to Mailman 3.3.2, import21 would add the entire 2.1 *_these_nonmembers and then delete the non-regexps. Possibly something went wrong in that process because of the truncation and left the new *_these_nonmembers in a broken state. Or possibly you just have a huge number of regexps in these list's accept_these_nonmembers. I suggest you use mysql to update this column's data for these three lists. to a value which unpickles as an empty list. This value is the python bytes b'\x80\x04\x5d\x94\x2e' (i.e. 5 bytes with hex values 80, 04, 5d, 94 and 2e). I'm not sure how that is stored in a MySQL BLOB, but I'm sure you can determine that. This will allow you to actually access those lists in Mailman. Also note that beginning with Postorius 1.3.3, the *_these_nonmembers settings are exposed on the List Settings -> Message Acceptance form. If you still have the original MM 2.1 config.pck files, you can examine them with Mailman 2.1's bin/dumpdb command to see what was in accept_these_nonmembers, and then see if the appropriate MM 3 nonmembers were created and if the remaining regexps are not too long, add them. If you need help with this, I will help. -- Mark Sapiro <mark(a)msapiro.net> The highway is for gamblers, San Francisco Bay Area, California better use your sense - B. Dylan _______________________________________________ Mailman-users mailing list -- mailman-users(a)mailman3.org To unsubscribe send an email to mailman-users-leave(a)mailman3.org https://nam02.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.mai…

4 years, 4 months

[MM3-users] Re: mailman 3 adding extraneous /var/ to pid path?

by Barry Warsaw

On Sep 21, 2017, at 18:35, Abhilash Raj <maxking(a)asynchronous.in> wrote: > > On Thu, Sep 21, 2017, at 02:42 PM, Mark Sapiro wrote: >> On 09/21/2017 01:29 PM, Andrew Hodder wrote: >>> >>> I keep getting around this by creating symlinks to the pid file and then >>> the stop works, but it's getting silly. It seems like mailman just assumes >>> CWD + /var/ + master.pid must be the location. Once I realized this, I >>> found I can do "mailman stop" so long as I'm only in /opt/mailman/ (since >>> it'd then add /var/master.pid). But that seems a little nonsensical. >> >> That is the normal expected behavior if devmode is enabled. Check your >> configs for this. > > I don't think `devmode` has anything to do with this, it only disables > sending out emails by accident and routes all emails to a pre-defined > address. Haven't looked at the relevant portion of the source so I might > be wrong. That’s correct. The [paths.*] sections are what control the file system layout, and [mailman]layout is what defines which of the [paths.*] sections to use. Cheers, -Barry

8 years, 2 months

[MM3-users] Re: Apache+mod_wsgi issue

by Guillermo Hernandez (Oldno7)

I top post just for confirming that all your recommendations worked well and I am aplying them in the rest of servers Thanks a lot and all of you have a happy new year. El 27/12/22 a las 9:23, Guillermo Hernandez (Oldno7) via Mailman-users escribió: > > El 26/12/22 a las 23:57, Mark Sapiro escribió: >> ... >>> I don't want to put more noise in the list. But I'm confused because >>> I always updated via pip the apache mod_wsgi thinking is was needed >>> for my configuration. If it is not, I will stop updating it. >> >> Based on the above you ARE using Apache mod_wsgi and if you are still >> also running uwsgi as >> >> /usr/local/bin/uwsgi --ini /usr/local/mailman3/uwsgi.ini & >> >> that is unnecessary. > > It clarifies a pair of things. Thanks a lot. > > I'll try deactivating that in one of my servers (the "quieter" one) > and see how it goes. > > Again, thanks for your kind support. > > > > -- ___________________________________________ Mailman's content filtering has removed the following MIME parts from this message. Content-Type: image/png Name: firma-GHP-emails.png Replaced multipart/alternative part with first alternative.

2 years, 11 months

[MM3-users] Re: MM3 Docker

by Odhiambo Washington

On Wed, 9 Jan 2019 at 19:25, Mark Sapiro <mark(a)msapiro.net> wrote: > On 1/8/19 10:50 PM, Odhiambo Washington wrote: > > > Now that has clarified the matter for me. That even when I am logged out, > > the scripts do the neccessary > > to activate the venv to run things. > > > As mentioned before, a virtualenv is not like a container. It is only a > directory that contains among other things a bin/ directory with > activate scripts and easy_install, python, pip and wheel commands. > > When you activate a virtualenv in your shell, all it really does is add > a deactivate command, modify the command prompt and insert its bin/ > directory at the beginning of its PATH. > > What this means is whether or not a virtualenv is "active", running a > command from the bin/ directory in the virtualenv is using the virtualenv. > Thank you very much for the elaboration. What used to confuse me was the early stage documents which detailed how to run MM3 in a virtual environment - I am talking about the BETA days. I think from this point onward, I have the complete foundation information to move forward. -- Best regards, Odhiambo WASHINGTON, Nairobi,KE +254 7 3200 0004/+254 7 2274 3223 "Oh, the cruft.", grep ^[^#] :-)

6 years, 11 months

[MM3-users] Re: Moderate mails from owner

by Malcolm Austen

On Wed, 20 Mar 2019 18:42:10 -0000, Mark Sapiro <mark(a)msapiro.net> wrote: > However, this is apparently not relevant because both the Member and > Non-Member actions are Hold. Thus, the post should be held unless it has > an Approved: header with the list's moderator_password attribute if > there is one, or it matches a Header filter rule with an accept action. At the risk of getting this really wrong - I have only limited MM3 knowledge - I will ask what is, I fancy, a closely related question ... My only experience of MM3 is with Ancestry's 'RootsWeb' installation. From the emails I can see a header "X-Mailman-Version: 3.1.0" but I can't find anything to identify Postorius which is the only interface I have to the system. From symptoms seen and described by others, I have deduced that the Postorius admin-per-user moderation options of "Default processing" and "List default" mean that a user may be set (by default on the rootsweb system!) to use the system default moderation settings rather than the list default settings. Are my deductions correct? And if so, could this be the explanation of the issue detailed in this thread? Apologies if I'm barking up entirely the wrong tree, Malcolm. -- Malcolm Austen <malcolm.austen(a)weald.org.uk>

6 years, 8 months

[MM3-users] Re: Hyperkitty shows "<" as <

by Andreas Schamanek

Thanks Mark, Your support is unparalleled. On Tue, 4 Apr 2023, at 10:44, Mark Sapiro wrote: > On 4/4/23 09:53, Andreas Schamanek wrote: >> All "<" and ">" in old and new messages are shown as "<" and ">" in >> message views _however_ they are shown correctly in overview abstracts e.g. >> at //server/mm3/hyperkitty/list/testlist@server/latest > > I don't think I've ever seen this. Certainly it is not the case on this list > with either current messages such as those in this thread > <https://lists.mailman3.org/archives/list/mailman-users@mailman3.org/thread/…> I tried to dig deeper and realized that on my test list "<" and ">" are only mangled when they surround email addresses as e.g. in <user(a)example.net> (ie. < user @ example . net >) I see that lists.mailman3.org renders such addresses as e.g. <a target="_blank" href="mailto:user@example.net">user(a)example.net</a> whereas my Hyperkitty shows them as <user(a)example.nett> But I have not yet found the responsible code or setting since it seems this is related to the option to obscure addresses. Or is my Hyperkitty 1.3.4 simply too old? -- -- Andreas :-)

2 years, 8 months

[MM3-users] Re: Mailman stops processing....... where to start looking for root cause?

by Stephen J. Turnbull

Mark Sapiro writes: > One of the runners has died, most likely the `out` runner, but > possibly another. The missing messages are stuck in a queue. Do ``` > ls -laR /opt/mailman/core/var/queue ``` to see in which queue the > messages are. The runner that processes that queue is gone or > possibly stuck. I have seen "stuck", in a multisite installation where Mailman's smarthost ended up being about 1000 miles from Mailman. I'm not sure what the issue is (we fixed the DNS, not Mailman :-), but every once in the while the connection would hang. In that case nothing showed up in the log. The only symptoms were the buildup in the out queue and a stale connection to port 25 on the smarthost. If your smarthost is not "network far" from the Mailman host, this is quite unlikely from my experience, although Mark may know better. In any case, I think the best way to start diagnosis is to check the Mailman and MTA queues for stuck messages. Of course in normal operation you will often see messages in the queue, but long queues or messages older than a minute are 99% a dead or stuck queue runner. Steve

1 year, 8 months

[MM3-users] Re: Disable Sign-up per domain

by Mark

On 2025-07-31 03:31, Mark Sapiro wrote: > Anything is possible, it's only code ;) > > You could make your own user_adapter.py which is modified like > > and set that as the ACCOUNT_ADAPTER. This depends on whatever is > proxying to Django setting the appropriate HTTP_HOST header. Thank you Mark. It's so close. After wrangling your solution a bit it's working a treat on my test server (wrangled solution below) . When I add it to the production server (same OS, MM installation, etc) I'm getting a "Bad gateway" error page and the log reports ... ... raise self.model.DoesNotExist( django.contrib.sites.models.Site.DoesNotExist: Site matching query does not exist. I cannot for the life of me get my head around that. I suspect it's something to do with what you said -- "whatever is proxying to Django setting the appropriate HTTP_HOST header." -- which I don't understand. Any additional advice gratefully received. ================================== /opt/mailman/venv/lib/python3.11/site-packages/django_mailman3/views/user_adapter.py ================================== COMMENT OUT THESE 2 LINES... # def is_open_for_signup(self, req): # return False REPLACE WITH THE FOLLOWING... def is_open_for_signup(self, req): if req.META['HTTP_HOST'] == 'lists.mydomain.com': return False else: return True ================================== /etc/mailman3/settings.py ================================== Append the following ... ACCOUNT_ADAPTER = 'django_mailman3.views.user_adapter.DisableSignupAdapter' ================================== systemctl restart mailman3 systemctl restart mailmanweb ==================================

4 months, 1 week

[MM3-users] Re: UnicodeEncodeError: 'ascii' codec can't encode character

by Marvin Gülker

Hi, Am 13. Dezember 2019 um 17:10 Uhr -0800 schrieb Mark Sapiro: >key is the <message-id> of the request message >request_type is 1 for a held message >mailing_list_id is a numeric index into the mailinglist table > >The issue is there are still held message requests there for the >messages you removed. You need to delete those entries from the _request >table. Thank you again for your valuable help. I've now taken the easy route and ran DELETE FROM _requests WHERE request_type=1; , i.e. I cleared the entire held messages queue. As it turned out, there were no requests of any other type than type 1. I've deleted all remaining files below /var/lib/mailman3/messages as well (after seeing it was only spam anyway, this time non-broken spam, but still spam). Now, finally, I can go to the held messages queue in Postorius again without a 500 error. The list of held messages there is now empty, but this is obviously expected. Let's hope that disabling SMTPUTF8 prevents the problem from further occuring. In any case, I've asked a friend report the problem to the Ubuntu tracker: https://bugs.launchpad.net/ubuntu/+source/mailman3/+bug/1856181 I'm back on track now. Thanks everyone! Marvin -- Blog: https://mg.guelker.eu

5 years, 11 months

[MM3-users] Re: Installing MM3 via pip fails…

by Patrick Ben Koetter

* Mark Sapiro <mark(a)msapiro.net>: > On 2/18/20 10:45 AM, Patrick Ben Koetter wrote: > > Has anyone managed to install MM3 recently using pip on Ubuntu 18? > > > > I followed the instructions at > > https://docs.mailman3.org/en/latest/install/virtualenv.html#virtualenv-inst…. > > As soon as I get to the Django part things get hairy. It seems Django requires > > something that was removed from it. > > > What exactly is the issue? I just tried following the doc starting at > <https://docs.mailman3.org/en/latest/install/virtualenv.html#virtualenv-setup> > and doing everything through > > (venv)$ python3 manage.py runserver > > and I encountered no issues. One possibility is the Ubuntu 18.04 box on > which I did this already had ruby-sass and python3-dev globally installed. > > sudo apt install ruby-sass > sudo apt install python3-dev > > Did you install these? Yes, I did. But to be sure I'll try on a vanilla VM and see how it goes if I install ruby-sass and python3-dev beforehand. p@rick -- [*] sys4 AG https://sys4.de, +49 (89) 30 90 46 64 Schleißheimer Straße 26/MG,80333 München Sitz der Gesellschaft: München, Amtsgericht München: HRB 199263 Vorstand: Patrick Ben Koetter, Marc Schiffbauer, Wolfgang Stief Aufsichtsratsvorsitzender: Florian Kirstein

5 years, 9 months

Search results for query "sapiro"