[OpenAFS] Mail delivery into OpenAFS

Nathan Ward nward@esphion.com
Mon, 07 Jul 2003 15:40:20 +1200


On Sun, 06 Jul 2003 19:54:57 -0700, Russ Allbery <rra@stanford.edu> wrote:

> Nathan Ward <nward@esphion.com> writes:
>
> There's no way to test first that doesn't involve a race condition.
>
Granted, but it at least shortens the window. Checking in multiple places 
should shorten this further, no?


> It's your mail, so if you can convince yourself that it's safe enough for
> you, that's your call.  I tend to avoid trying to make up and prove
> properties of locking methods and just use established, well-regarded 
> ones instead; it saves wear and tear on my brain.

On Sun, 6 Jul 2003 23:14:22 -0400 (EDT), Stephen Joyce 
<stephen@physics.unc.edu> wrote:

> PS. Any specific reason you're reinventing the wheel rather than use a
> proven email solution?  (I personally like cyrus--while I never got it to
> compile with AFS support, I found I don't really need it since it has
> built-in acl and quota support).

All home directories are on AFS.
Users want mail in thier home directories. I want user's mail quotas as 
part of thier overall disk quota. Having a second quota for mail is all 
well and good, until the user recieves a large file (this is a common thing 
in our company) and it takes the whole thing. At least with a single large 
quota of thier home directory they can free up some space and recieve the 
mail in the unlikely event that it bounces. It is unlikely that any user 
will see thier AFS quota full anytime soon. It is perhaps worth noting that 
we have alot more disks dedicated to AFS than to the mail server.