[Mailman-Users] Test messages

[Adam McGreggor]

On Thu, Jul 29, 2010 at 12:12:52PM +0100, Steff Watkins wrote:
> One thought/idea I've had is to have a "test" list which has only my
> work email address, an offsite email address and maybe one or two of my
> colleagues (for verification purposes), lock it right down so that only
> the members can use it, add the "Me too" tag so that we'd receive any
> emails we sent to the list and use that to check that the mailman
> service is functional. 

That might be one way; although useful for those of us not on some
tuppence ha'penny shoe-string "ISP" who believe in limiting the amount
of mail transmissions (and where the punter may not know of those
limits).

Of course, MTA problems may also exist, which that approach wouldn't
examine (nor do I see mention of testing the web interfaces work).

*brainfart*
Perhaps a Mailman-admin's 'test suite' might be useful? And
presumably have the option to run (gah!) via the web-admin interface,
as well as the command line

> Does that sound a valid way of doing it, too much
> effort FWIW or missing the point as it wouldn't test a particular list
> just the mailman service?

using config_list on a "troublesome" list, a spot of sed/withlist, and
config_list on a new list, may be one way to transfer the list
settings over; possibly followed by remove_members -a and add_members;
all of that could be made into a two arguements "clonelist" script.

This doesn't, of course, help, if Mailman's not running -- never rule
out (or forget about) the obvious -- nor, say if there's a lack of
disk-space.

I can't remember what, if anything, was decided about test lists in
MM3. 

I'll be waiting for MM3, before I start fiddling with trending Mailman
in Nagios; for now, my "last posts" script/output does the job well
enough (and I am a lazy sysadmin) for me.

-- 
``What is a committee?  A group of the unwilling,
  picked from the unfit, to do the unnecessary.''  (Richard Harkness)