Re: MySQL Reengineering Project

[Sergei Golubchik]

Hi, Mikiya!

On May 18, Mikiya Okuno wrote:
> Hi,
> 
> Does this include a large pages support?

No, it does not.
May be later.
 
> I encountered a problem that the performance went down when several
> mega bytes were assigned for each read buffer, read rnd buffer and
> sort buffer on a Nocona system. I suspect that many TLB cache miss
> were happening at that time. Currently, MySQL only supports large
> pages for InnoDB. If MySQL will have a slab allocator which supports
> large pages like memcached, the performance will improve due to less
> TLB cache miss, won't it?

large pages are used for myisam keycache, that's why the support for
them was added in the first place.

There's no problem to use them elsewhere, where it make sense.

But there aren't many places where it make sense, as, indeed, we don't
have a generic purpose allocator in the server.

Regards / Mit vielen Grüßen,
Sergei

-- 
   __  ___     ___ ____  __
  /  |/  /_ __/ __/ __ \/ /   Sergei Golubchik <serg@stripped>
 / /|_/ / // /\ \/ /_/ / /__  Principal Software Engineer/Server Architect
/_/  /_/\_, /___/\___\_\___/  Sun Microsystems GmbH, HRB München 161028
       <___/                  Sonnenallee 1, 85551 Kirchheim-Heimstetten
Geschäftsführer: Thomas Schroeder, Wolfgang Engels, Wolf Frenkel
Vorsitzender des Aufsichtsrates: Martin Häring