Re: Better directory hashing for cyrus-imapd-2.0.x

2001-03-11 Thread Lawrence Greenfield
It isn't configurable. This hasn't been an issue for us yet and thus we haven't bothered writing the code to do so. I generally try to recommend to people to look at different filesystems if they're very concerned about this. Larry --On Monday, January 15, 2001 03:48:51 PM +0100 Stephane Ent

Re: Better directory hashing for cyrus-imapd-2.0.x

2001-01-15 Thread Stephane Enten
On Fri, Jan 12, 2001 at 11:43:59AM -0600, [EMAIL PROTECTED] wrote: > Larry Greenfield writes: > > > >This sort of directory hash is very easy to implement in Cyrus; a > >better directory hash than the one included wasn't chosen since we > >wanted something simple that worked well enough for u

Re: Better directory hashing for cyrus-imapd-2.0.x

2001-01-12 Thread mills
Larry Greenfield writes: > >This sort of directory hash is very easy to implement in Cyrus; a >better directory hash than the one included wasn't chosen since we >wanted something simple that worked well enough for us. > >Unfortunately, due to extreme laziness on my part, the directory hash >funct

Re: Better directory hashing for cyrus-imapd-2.0.x

2000-12-05 Thread Lawrence Greenfield
This sort of directory hash is very easy to implement in Cyrus; a better directory hash than the one included wasn't chosen since we wanted something simple that worked well enough for us. Unfortunately, due to extreme laziness on my part, the directory hash function used is used in multiple plac

Better directory hashing for cyrus-imapd-2.0.x

2000-11-27 Thread mills
I'm in the process of setting up a cyrus server for about 40,000 users. Of these, about 3/4 have userids that begin with the string `um', followed by an abbreviation of the person's last name, possibly followed by a 1 to 3 digit number. The rest have userids derived from their last names and init