number of recipients with LMTP

[Stephen L. Ulmer]

 "P.A.Chambers" == Phil Chambers <P.A.Chambers at exeter.ac.uk> writes:

  P.A.Chambers> To get the maximum filespace efficiency out of cyrus
  P.A.Chambers> would need the limit to be high enough to allow one
  P.A.Chambers> message to all of out 20,000 users in one go.  That
  P.A.Chambers> would result in one copy of the message and 19,999
  P.A.Chambers> links.  Would there be any adverse consequences, from
  P.A.Chambers> the cyrus point of view, of telling exim to allow
  P.A.Chambers> 20,000 recipients in a single transfer?

The reason *we* don't do that: If you have to restore the filesystem
from backup. Depending on your backup technology, make sure that all
of those links won't become separate files (ours would).

We were unwilling to accept the risk of being unable to restore our
backup to a disk plant of the size it was taken from.  We also
couldn't come up with an easy (read inexpensive) way to calculate the
expansion factor.

Good Luck,

--
Stephen L. Ulmer                              ulmer at nersp.nerdc.ufl.edu
Senior Systems Programmer                         http://www.ulmer.org/
Northeast Regional Data Center                      VOX: (352) 392-2061
University of Florida                               FAX: (352) 392-9440