Re: Linux kernel turning for mail performance?

[cmikk]

On Thu, 3 Feb 2000 13:29:03 -0700 , [EMAIL PROTECTED] writes:
> On Thu, Feb 03, 2000 at 08:51:17PM +0100, Andre Oppermann wrote:
> >  2. *BSD FFS (with and without softupdates) honors fsync() for data
> >     AND metadata in any case.
> > 
> >     -> The fsync() call returns when the data and metadata is written
> >        to the disk.
> > 
> >     == reliable for DJB
> 
> Why does he say not to use softupdates then?
> 
> I guess the main point is that any filesystem that honors fsync()
> for data and metadata is okay.

There's the rename() issue also: if rename() [ or
link() ] return before the modified directories have
been committed, then qmail-local will drop the mail,
when delivering to a maildir.  qmail-queue will
similarly drop mail, if it is told falsely that it
has renamed intd/whatever to todo/msgnum.

I don't know if McKusick's softupdates code will
let rename() return earlier than documented.
 
-- 
Chris Mikkelson  | ... a pet peeve of mine is people who directly edit 
[EMAIL PROTECTED] | the .cf file instead of using the m4 configuration
                 | files ... I treat the .cf file as a binary file
                 | - you should too.  --- Eric Allman