On Tuesday 23 January 2001 15:31, Russell E Glaue wrote:
>This sounds interesting, especially with Linux 2.4 which supports more
>processors. Is it being said that MySQL was not written to take advantage
>of SMP, or is MySQL written to rely on the OS for any multi-processor
>capabilities? Would it really be difficult to code SMP support into MySQL?
>-RG
MySQL can take advantage of SMP, but not withing one thread ( connection). So
one query will always be run on one CPU at a time. However, if you have
parallel queries on several connections, they can be on different CPUs at the
same time. Mauricio, as I understood, was talking about distributing across
several machines. In that case, the best thing to do is replication. There is
a possiblity to get really good performance with several machines accessing
the same data directory residing on the same disk array through GFS, as long
as the file locking issues could be dealt with graciously, but this is only
theory - we have not tried it. On the other hand, replication is a tried way
to scale read performance.
--
MySQL Development Team
__ ___ ___ ____ __
/ |/ /_ __/ __/ __ \/ / Sasha Pachev <sasha@stripped>
/ /|_/ / // /\ \/ /_/ / /__ MySQL AB, http://www.mysql.com/
/_/ /_/\_, /___/\___\_\___/ Provo, Utah, USA
<___/
