Timeout on visiting pending moderator requests
I have a netbsd 1.6 machine running Mailman 2.1.5 serving about twenty lists. One of my lists has a problem where it's pending moderator requests page can't be viewed, it just times out. I'm speaking about the URL that ends in admindb/<listname>. I see the vette log increase by about 320Kb everytime I visit the pending moderator requests page. There are 629 files named data/heldmsg-<listname>-* using 3100Kb of disk space. I haven't been able to decipher the problem.
The farthest I've gotten is to run 'dumpdb pending.pck' for that list. I can't make sense of the output, but if it would be useful in debugging the problem I could mail it to this list or one of you as an attachment. I can individually reject messages via the email-to-reject reply mechanism for held messages.
This is the contents of the lists/<listname> directory:
-rw-rw-r-- 1 morgan mailman 1706 Apr 13 2002 admindbpreamble.html -rw-rw---- 1 morgan mailman 47410 Jul 26 06:05 config.db -rw-rw---- 1 morgan mailman 56959 Jul 26 06:05 config.db.last -rw-rw---- 1 nobody mailman 53170 Dec 28 01:12 config.pck -rw-rw---- 1 mailman mailman 53170 Dec 28 01:11 config.pck.last -rw-rw---- 1 mailman mailman 14402 Dec 28 00:51 digest.mbox -rw-rw-r-- 1 morgan mailman 189 Apr 13 2002 handle_opts.html -rw-rw-r-- 1 morgan mailman 1026 Apr 13 2002 headfoot.html -rw-rw-r-- 1 morgan mailman 3136 Apr 13 2002 listinfo.html -rw-rw-r-- 1 morgan mailman 2961 Jul 25 16:59 next-digest -rw-rw-r-- 1 morgan mailman 49 Jul 25 16:59 next-digest-topics -rw-rw-r-- 1 morgan mailman 4106 Apr 13 2002 options.html -rw-rw---- 1 mailman mailman 7304 Dec 28 01:11 pending.pck -rw-rw-r-- 1 morgan mailman 547352 Jul 26 06:04 request.db -rw-rw-r-- 1 mailman mailman 2975275 Dec 28 01:11 request.pck -rw-rw-r-- 1 morgan mailman 1169 Apr 13 2002 roster.html -rw-rw-r-- 1 morgan mailman 198 Apr 13 2002 subscribe.html
I've run 'check_perms' and 'check_db' and everything checks out OK.
Got anything for me?
Morgan
Morgan Fletcher, morgan@hahaha.org Oakland, CA
Hi,
Mailman 2.1.6 will discard old pending moderator requests automatically. So, I you don't have important pending messages to be approved, wait for the next release ;-) Or, you can use bin/discard to discard old pending requests like:
% cd <prefix> % find data -ctime +7 -name heldmsg\*.pck | xargs bin/discard
Cheers,
Morgan Fletcher wrote:
I have a netbsd 1.6 machine running Mailman 2.1.5 serving about twenty lists. One of my lists has a problem where it's pending moderator requests page can't be viewed, it just times out.
-- Tokio Kikuchi, tkikuchi@ is.kochi-u.ac.jp http://weather.is.kochi-u.ac.jp/
Tokio Kikuchi <tkikuchi@is.kochi-u.ac.jp> writes:
% cd <prefix> % find data -ctime +7 -name heldmsg\*.pck | xargs bin/discard
That works but the problem remains after removing some of the heldmsg files for that list - the admindb/listname URL never loads. There are now 161 of the heldmsg-<listname>-*.pck files.
I can load the pending administrative requests pages for all my other lists, including those with bigger admin queues.
I appreciate your help,
Morgan
Morgan Fletcher, morgan@hahaha.org Oakland, CA
Morgan Fletcher <morgan@hahaha.org> writes:
I have a netbsd 1.6 machine running Mailman 2.1.5 serving about twenty lists. One of my lists has a problem where it's pending moderator requests page can't be viewed, it just times out. I'm speaking about the URL that ends in admindb/<listname>. I see the vette log increase by about 320Kb everytime I visit the pending moderator requests page. There are 629 files named data/heldmsg-<listname>-* using 3100Kb of disk space. I haven't been able to decipher the problem.
Is there anything I can do to debug this problem? I can use the discard tool to discard messages blindly, but I would like to again be able to use the admindb/<listname> URL to review and act upon the admin queue.
Morgan
Morgan Fletcher, morgan@hahaha.org Oakland, CA
participants (2)
-
Morgan Fletcher -
Tokio Kikuchi