Matthias Andree wrote:
Joerg Schilling <[EMAIL PROTECTED]> writes:
Bill Davidsen <[EMAIL PROTECTED]> wrote:
This sounds a bit confused. Are you able to describe your concern?
You said star doesn't run as fast on Linux as Solaris, you can probably
fix that problem by using O_DIRECT, so the writes and reads don't
compete for the same buffer memory.
And I already mentioned that it is higly improbable that O_DIRECT will
speed up on Linux. A better buffer cache (this means improving Linux) helps
much more than avoiding it.
How would a buffer cache improve the situation for star? I may have
missed the beginning of the discussion, but caching
write-once-then-forget data seems pointless.
I think he means one which didn't do that useless caching, although most
improvements also bite in some cases. I had a patch in which assumed
that if a program wrote more than N bytes without a read or seek that
the data should be sent to the drive NOW. Eliminated writing a full CD
to buffer, with data from cache, then closing the file and having the
drive go dead busy for a minute.
--
bill davidsen <[EMAIL PROTECTED]>
CTO TMR Associates, Inc
Doing interesting things with small computers since 1979
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
