On Fri, 2006-05-12 at 10:59 -0400, RV Tec wrote:
> Yeah, I'm aware of that. Since glibc 2.4 has only NPTL, I was
> wondering if it
> is possible to MySQL use a threading system of its own. However, what
> I want to
> know, is a way to confirm that it has been compiled against NPTL.
>
> This appe
I'll add "make sure logs and data are on separate partitions" so
you're not doing excessive seeking back and forth.
-Sheeri
On 5/12/06, Dan Buettner <[EMAIL PROTECTED]> wrote:
Hope it is useful.
I agree, you may want to look at adding another card and disks, for
speed and to segregate the vari
Hope it is useful.
I agree, you may want to look at adding another card and disks, for
speed and to segregate the various operations (temp, logging, data).
Splitting up your MYD and MYI files may help, though if you have enough
RAM to keep indexes in memory, maybe you don't need to do that.
On 5/12/06, RV Tec <[EMAIL PROTECTED]> wrote:
Buettner,
First of all, thanks a lot for your reply!
This server has 4 disks to MySQL usage, in two pairs of RAID-1, connected
to a single channel (ok, I realize now this means a bottleneck) LSI
PCIe card.
One RAID1 for MySQL logging and temp space
Good morning RV -
On your 3rd question, about how to make things faster:
More RAM should help by allowing the server to keep more/all of the
indexes in memory, enabling much faster access. Be sure to adjust the
cache settings in your my.cnf file after adding RAM. (Keep in mind -
some my.cnf
1) Is there a way to see MySQL using both processors? Is SMP helpful in
this case? (This server is dedicated to MySQL, only one instance).
PS (*nix) should tell you how your processors are used.
PS does show me about the CPU usage, but it doesnt tell me which
processor, or if they're being use
Buettner,
First of all, thanks a lot for your reply!
This server has 4 disks to MySQL usage, in two pairs of RAID-1, connected
to a single channel (ok, I realize now this means a bottleneck) LSI
PCIe card.
One RAID1 for MySQL logging and temp space, and the other pair for the
database files
RV Tec schrieb:
Folks,
I had some recommendations about operating system last time I posted,
and decided to follow it. It's been a couple of weeks running Gentoo
Linux 2006.0 SMP 2.6.15-gentoo-r5, with glibc 2.4 (NPTL), gcc 3.4.4, XFS
as my FS, deadline scheduler and this has proven to be rea
Folks,
I had some recommendations about operating system last time I posted, and
decided to follow it. It's been a couple of weeks running Gentoo Linux
2006.0 SMP 2.6.15-gentoo-r5, with glibc 2.4 (NPTL), gcc 3.4.4, XFS as my
FS, deadline scheduler and this has proven to be really stable -- MyS
Hi again. :-)
On Thu 2002-09-05 at 14:18:10 -0500, [EMAIL PROTECTED] wrote:
[...]
> 3) I'm somewhat at a loss for this one and perhaps the answer is more
> obvious than not. I have 257 total tables from my main DB and mysql. I
> figured this by a "ls -al var/ | grep -c MYD". How can I possibly
First of all, I forgive the rather lengthy post.
Thanks for the repl(y|ies) Benjamin. Decreasing the key_buffer should
be my first step. Back to the questions:
3) I'm somewhat at a loss for this one and perhaps the answer is more
obvious than not. I have 257 total tables from my main DB and
Benjamin Pflugmann wrote:
>OTOH, if this is a MySQL-only machine, 3GB are plenty and 100MB more
>or less used do not really matter (regarding free memory), so I would
>simply set it to use about 400MB are forget about it.
>
>
Remember to actually benchmark your differences too if possible (with
Hi.
On Thu 2002-09-05 at 09:09:07 -0500, [EMAIL PROTECTED] wrote:
> For clarity sake, assume the following:
>
> Red Hat Linux 7.1
> 2.4.8 kernel
> MySQL 3.23.42
> MyISAM databases
> 3GB RAM
> P3/700 x 4
> 15GB database spanned across ~200 tables
>
> Key_reads / Key_read_request = 0.00059875
Rummaging through some docs on performance and have come up with some
questions. Let me preface by saying, we don't have any performance
problems. I inherited this monster of a database and am running through
the configuration to make sure that it is indeed setup for optimum
performance.
For cl
14 matches
Mail list logo
