We still have a problem with courier 1.2.3 / multiple mail deletes, with
no NFS involved. There is a hardware RAID array on the machine holding the
mail, that also runs the email server...
On Wed, 6 Dec 2000, Brian Candler wrote:
On Tue, Dec 05, 2000 at 08:39:41PM -0500, Sam Varshavchik wrote:
Probably not related, but I currently have a problem with sqwebmail where
messages which are deleted remain in the folder listing. If you manually
delete the sqwebmail-curcache.db1 file it corrects itself. Haven't had a
chance to try and nail it down yet.
This usually indicates that you are using NFS, or a reasonable facsimile,
and the clocks on the NFS server and client are not synchronized.
This is running under FreeBSD-4.1.1 with the maildrops stored on a NetApp.
See, I was right.
Get ntp running on both boxes, and have them talk to each other.
Hmm. How precise does the time synchronisation need to be? I need to go
check the current status, but a couple of years ago Netapps didn't support
ntp. I had to set up a cronjob to 'rsh date ...' to reset the clock
periodically. You can probably count on it being within a second or two.
6 messages in net.sourceforge.lists.courier-usersRe: [courier-users] Multiple email de...
