[Mailman-Users] mailman gets stuck, stops sending messages

[Tokio Kikuchi]

Hi, sorry for my late response

David Relson wrote:

> I run several low traffic lists (rarely more than 10 messages per day
> and fewer than 250 users per list).  I'd worry about a restart while
> mailman is in operation.  Not knowing the code, there _could_ be
> problems in nuking it unexpectedly.  With my low volume, "unexpected"
> isn't likely to cause damage, but I'd be careful.

Are you also running debian patched mailman? I found a funny code in the 
debian mailman patch 
(http://ftp.debian.org/debian/pool/main/m/mailman/mailman_2.1.5-3.diff.gz)

+--- mailman-2.1.4.orig/bin/mailmanctl
++++ mailman-2.1.4/bin/mailmanctl
+@@ -415,6 +414,13 @@
+         # won't be opening any terminal devices, don't do the 
ultra-paranoid
+         # suggestion of doing a second fork after the setsid() call.
+         os.setsid()
++
++        # Be sure to close any open std{in,out,err}
++        devnull = os.open('/dev/null', 0)
++        os.dup2(devnull, 0)
++        os.dup2(devnull, 1)
++        os.dup2(devnull, 2)
++
+         # Instead of cd'ing to root, cd to the Mailman installation home
+         os.chdir(mm_cfg.PREFIX)
+         # Clear our file mode creation umask

I wonder if this is needed at all and it may cause trouble because 
stdout and stderr (1, 2) are set O_RDONLY.

Mailmanctl is a watchdog for qrunners and it is funny if mailmanctl is 
also needed a watchdog. In other installations such as Solaris and 
FreeBSD I admin, such a patch is not needed. Try edit these lines and 
start mailmanctl again and feedback the packagers with devian if it 
works. (or, write this list again if it doesn't work. ;-)

-- 
Tokio Kikuchi, tkikuchi@ is.kochi-u.ac.jp
http://weather.is.kochi-u.ac.jp/