Mailman 3 March 2007 - Mailman-Users

Display error for 'members' page
by Rick Pasotto March 20, 2007

March 20, 2007

I'm running 2.1.9 under debian etch. I tried to clone a member from the command line (something I've done many times previously) but it failed with a NotAMemberError error. The traceback indicated that the *new* address was the problem. Well, of course the new address was not (yet) a member! Anyway, now when I try to use the web interface the 'members' page won't display. I get a 'Bug in Mailman' error. I have other lists on this machine (different domains) and they still work fine. How do I fix the problem? -- "No one can make you feel inferior without your consent." -- Eleanor Roosevelt Rick Pasotto rick(a)niof.net http://www.niof.net

2 5

How to move lists/subscribers to new machine?
by Peter March 20, 2007

March 20, 2007

I have an old installation (2.1.5) that I want to move to a new machine (running 2.1.8). I was wondering how to go about moving the lists and subscribers over. After that I will worry about the archives. The first machine is running Mandrake Linux and the new machine is running OpenBSD 4.0. Thanks to any responders, Pedro

2 1

Server migration question
by Steve Burling March 20, 2007

March 20, 2007

So, we pretty much successfully moved our mailman installation from the old Solaris box to the new Red Hat box. Only one small problem. Of course. We (and by we, I mean me) had failed to reduce the TTL for the MX record that points to our mail server, which is now on a new machine. So some mailman mail is still being delivered to the old machine. I've stopped the qrunners on that machine, so things are just piling up in the qfiles/in directory. >From my examination of the files there, 99% of what's coming in is SPAM, so I don't care about delivering it. But for the remaining 1%, I'd like to eventually deliver that mail, via the new mailman installation, so that it ends up archived in the correct place. Is there any reasonably-straightforward way that I can take the *.db and *.msg files that I care about and move them to the new installation? Or is what I really want to do to use bin/inject, passing it the individual .msg file? Thanks in advance for any hints... -- Steve Burling <mailto:srb@umich.edu> University of Michigan, ICPSR Voice: +1 734 615.3779 330 Packard Street FAX: +1 734 647.8700 Ann Arbor, MI 48104-2910

3 3

Which settings controls mailman email address disiplayed?
by Ryan Golhar March 20, 2007

March 20, 2007

Forgive me...I'm new to mailman and am having some trouble. I know the settings are in mm_cfg.py or data/sitelist.cfg but I can't seem to find where. I've set up mailman on a RedHat linux server, and created a mailing list. My first problem is on the mainpage (http://xxx.xyz.edu/mailman/listinfo) it says: If you are having trouble using the lists, please contact mailman(a)xxx.xyz.edu. How do I change it from mailman(a)xxx.xyz.edu to something else? Where is that setting stored? Ryan

2 3

Newsgroup interface - upload to mailman not working
by David Beaumont March 19, 2007

March 19, 2007

We interface to a newsgroup, details: "The hostname of the machine your news server is running on. news.annexcafe.com The name of the Usenet group to gateway to and/or from. annexcafe.dba.bargemain" Emails are transferred to the newsgroup OK. Until September last year newsgroup postings were picked up by mailman and distributed as emails. However since then that bit has stopped working, any ideas why? Our supplier says the newsrunner process is running. Thanks David

3 5

using "subscribe" in email subject to subscribe users?
by JT March 19, 2007

March 19, 2007

How do I configure mailman to accept "subscribe" email? Is there a way to configure my Mailman version 2.1.9.cp2 (or is it already configured) so that I can have subscribers email my list with "subscribe" in the subject or body? Thanks! -traynor

1 0

Re: [Mailman-Users] Mailman possible overloading Postfix
by Johann MacDonagh March 17, 2007

March 17, 2007

Did you have to make any special configuration changes? Johann MacDonagh Paul Tomblin wrote: > Quoting Johann MacDonagh (johann(a)macdonaghs.com): > >> Does anyone have any experience running Mailman on a low end server? I >> can double the specs on this VPS for an extra $10 a month. I'll probably >> test that out and see how well it works. >> > > I ran my mailman lists (with Postfix and Apache2 on Debian Sarge) on a > Linode 128 VPS for a few years with no problems. I had about 15 lists, > the biggest with about 150 members, and transmitted about 500 messages a > day. I never got throttled and never had problems with dozens of > simultaneous deliveries. > >

2 1

Mailman possible overloading Postfix
by Johann MacDonagh March 17, 2007

March 17, 2007

I'm setting up my Mailing list on a VPS (virtual private server) I have. It's set to have about 128MB of RAM. I'm running a simple Fedora Core 5 server. I've got Postfix / QRunner / Apache 2 running on it. I imported a mailing list that has about 80 subscribers. I attempted to post a message to all of them and this is what I got in my mail log. Mar 16 19:23:05 myhost postfix/master[23796]: warning: process /usr/libexec/postfix/bounce pid 3190 exit status 127 Mar 16 19:23:05 myhost postfix/master[23796]: warning: /usr/libexec/postfix/bounce: bad command startup -- throttling I used to get a lot of these throttling messages (especially when I held 3 or 4 messages in the mailman moderation queue and accepted them all at once. Since then I've attempted to slow down the speed at which it sends. Here's my postfix main.cf: smtpd_banner = $myhostname ESMTP $mail_name myhostname = mail.mydomain.org mydestination = mail.mydomain.org, localhost, localhost.localdomain mynetworks = 127.0.0.0/8 mailbox_size_limit = 0 recipient_delimiter = + inet_interfaces = all relay_domains = lists.mydomain.org canonical_maps = hash:/etc/postfix/canonical transport_maps = hash:/etc/postfix/transport smtpd_recipient_restrictions = permit_mynetworks, permit_sasl_authenticated, reject_unauth_destination, reject_unauth_pipelining smtpd_data_restrictions = reject_unauth_pipelining default_process_limit = 20 mailman_destination_recipient_limit = 1 And the important parts of my mm_cfg: DEFAULT_URL_HOST = 'lists.mydomain.org' DEFAULT_EMAIL_HOST = 'lists.mydomain.org' DEFAULT_URL_PATTERN = 'http://%s/' PRIVATE_ARCHIVE_URL = '/mailman/private' IMAGE_LOGOS = '/icons/' PUBLIC_ARCHIVE_URL = 'http://%(hostname)s/archive/%(listname)s/' MTA = None SMTP_MAX_RCPT = 20 # Because we've overriden the virtual hosts above add_virtualhost # MUST be called after they have been defined. add_virtualhost(DEFAULT_URL_HOST, DEFAULT_EMAIL_HOST) This seemed to fix it somewhat. I only got one error (the one I posted above) when sending one message. Here's the final strange thing. After a message gets delivered to all subscribers, the mailman site will be unavailable. Sometimes it says: "Oops, we've hit a bug in Mailman <undetermined version>". and sometimes it says something along the lines of: "Oops, we've hit a bug. .... The issue was very low level and a trace probably wouldn't help you." None of the mailman logs show anything regarding this. It seems like the server doesn't have enough resources to display the mailman page (and perhaps even fails to update the log because of a lack of resources). Does anyone have any experience running Mailman on a low end server? I can double the specs on this VPS for an extra $10 a month. I'll probably test that out and see how well it works. Thanks, Johann MacDonagh

3 5

mailman with fetchmail, ISP smtp, blocked port 80
by Nick Urbanik March 17, 2007

March 17, 2007

Dear Folks, I have just moved to a new ISP which blocks ports 80, 25 incoming. I have hosted (and hope to continue to host) two mailing lists, which worked fine till the change. I also have access to a web server pointed to by dns A record nicku.org on which I cannot install mailman, but on which I can install redirects, use mod_rewrite, and configure apache on. Postfix listens on port 25 on www3.nicku.org and can send mail via ISP mail server. An MX record for nicku.org points to mail.nicku.org, which in turn points as an A record to an email hosting service. www3.nicku.org is the new address of this machine hosting mailman. I would like to ask: 1. Any suggestions on a good way to configure mailman in this situation? My current method (dns points to this machine, www3.nicku.org, apache listening on port 8080, redirect from remote nicku.org to www3.nicku.org:8080) seems not right yet; attempts to click on the "Submit All Data" button on the admindb page just return the same page, without any apparent action having taken place. 2. I am running fetchmail as root, with a .fetchmailrc as shown below. Does that look correct? Any suggestions? 3. the /var/log/mailman/smtp-failure log shows multiple (failed) attempts to send mail to fetchmail-daemon(a)www3.nicku.org; how do I persuade mailman to send to the correct address? 4. Are there any special settings I need for mailman to match this arrangement better? Currently the changes to my mm_cfg.py are: DEFAULT_URL_HOST = 'www3.nicku.org:8080' DEFAULT_EMAIL_HOST = 'nicku.org' add_virtualhost(DEFAULT_URL_HOST, DEFAULT_EMAIL_HOST) Note that people access the web pages via http://nicku.org/mailman/listinfo/camwest-discuss and http://nicku.org/mailman/listinfo/camwest-announce. set daemon 180 set syslog set postmaster nicku set no bouncemail set no spambounce poll mail.nicku.org proto pop3 user user01 pass passwordxxx is nicku here user user02 pass passwordxxx is auser here user user03 pass passwordxxx is anotheruser here user user04 pass passwordxxx is mailman here user user05 pass passwordxxx is mailman-admin here user user06 pass passwordxxx is camwest-announce here user user07 pass passwordxxx is camwest-discuss here user user08 pass passwordxxx is camwest-announce-admin here user user09 pass passwordxxx is camwest-announce-bounces here user user10 pass passwordxxx is camwest-announce-confirm here user user11 pass passwordxxx is camwest-announce-join here user user12 pass passwordxxx is camwest-announce-leave here user user13 pass passwordxxx is camwest-announce-owner here user user14 pass passwordxxx is camwest-announce-request here user user15 pass passwordxxx is camwest-announce-subscribe here user user16 pass passwordxxx is camwest-announce-unsubscribe here user user17 pass passwordxxx is camwest-discuss-admin here user user18 pass passwordxxx is camwest-discuss-bounces here user user19 pass passwordxxx is camwest-discuss-confirm here user user20 pass passwordxxx is camwest-discuss-join here user user21 pass passwordxxx is camwest-discuss-leave here user user22 pass passwordxxx is camwest-discuss-owner here user user23 pass passwordxxx is camwest-discuss-request here user user24 pass passwordxxx is camwest-discuss-subscribe here user user25 pass passwordxxx is camwest-discuss-unsubscribe here where userxx is the username configured on the email hosting service, to which dns A record mail.nicku.org points. I will be grateful for any suggestions. -- Nick Urbanik http://nicku.org nicku(a)nicku.org GPG: 7FFA CDC7 5A77 0558 DC7A 790A 16DF EC5B BB9D 2C24 ID: BB9D2C24

2 1

how to re-activate large number of members?
by Ricardo Kleemann March 16, 2007

March 16, 2007

Hi, I have a couple of large lists (> 15,000 members) in which a significant percentage of them have been deactivated. I'm not sure why, I don't think that all of them are really consistently bouncing. I wanted to reactivate all of them and let mailman take its course once again in processing the bounces. But it's impossible to do it individually. Is there a way to change this en masse? Ricardo

2 1