I may not be understanding your issue correctly here, but the log buffer
flush is not tied to the block transfer through the interconnect AFAIK. The
log buffer flush is an independent event.
--
HTH. Additions and corrections welcome.
Pete
Author of "Oracle8i: Architecture and Administration Exam Cram"
"Controlling developers is like herding cats."
Kevin Loney, Oracle DBA Handbook
"Oh no, it's not. It's much harder than that!"
Bruce Pihlamae, long-term Oracle DBA
"Slava" <leichivp_at_my-deja.com> wrote in message
news:a775741a.0201310716.68dc5e46_at_posting.google.com...
> "Howard J. Rogers" <dba_at_hjrdba.com> wrote in message
news:<3c584313$0$8456$afc38c87_at_news.optusnet.com.au>...
> > One word (OK, two then!): cache fusion. Transferring a block from one
node
> > to another is now around 100 times faster than it was in OPS, because
there
> > is no block pinging to disk.
>
> Hi,
> probably your statement based upon the same diagram which I found in
> Mai Cutler's article "Oracle9i Real Application Clusters on HP" at
> Oracle OpenWorld.
> According to the article, block transfer from remote cache took 100ms
> in Oracle 8i write mode due to ping to disk and the same operation
> takes only 1 ms in
> Oracle 9i read and write mode because of improvement of cache fusion
> phase II.
> It seems very strange that no difference between times in read and
> write mode
> in 9i because as I know in write mode instance should flush log buffer
> to disk
> before block transfer to the remote cache.
>
> Regards,
> Slava.
Received on Thu Jan 31 2002 - 10:01:32 CST
