List:General Discussion« Previous MessageNext Message »
From:Sasha Pachev Date:July 15 1999 4:48am
Subject:Re: Optimizing a P133 for MySQL
View as plain text  
Van wrote:
> 
> Hey Folx, I'm finally getting an MySQL server on a T-1.  I'm recompiling
> the kernel as we speak on a Slakware 4.0 box.  It's characteristics
> follow:
> dude:~# cat /proc/cpuinfo
> processor : 0
> vendor_id : GenuineIntel
> cpu family : 5
> model : 2
> model name :
> Pentium 75 - 200
> stepping : 12
> cpu MHz : 132.955642
> fdiv_bug : no
> hlt_bug : no
> sep_bug : no
> f00f_bug : yes
> fpu : yes
> fpu_exception : yes
> cpuid level : 1
> wp : yes
> flags : fpu vme de pse tsc msr mce cx8
> bogomips        : 53.04
> 
> I'm running this on a T-1 with a 3c509x, or wd8013 card (whichever works)
> and have 48 Meg SIMMS 40ns NP ram and 56 Meg of swap set up on a single
> samsung 2 Gig hard-drive.
> 
> The application is for many clients, but, the biggest is a headhunter in a
> specific geographic area who's going to be pairing companies w/
> candidates, and the machine's job is to allow for uploads of resumes and
> searches by both candidate and company on job openings and candidate
> skills, respectively.
> 
> My questions are:
> Are there any specific 2.2.X Linux kernel tweakings I can implement during
> the kernel build to ensure optimal performance w/ these hard-ware
> settings?
> 
> Are there specific compiler flags I should throw the mysql 3.22.24 build?
> 
> I'm compiling in apache 1.3.6b with mod_auth_mysql, php_3.07, and probably
> gd for graphical hit reporting.
> 
> There will be 15-30K pages with 150 httpd processes to start, and, I'm not
> expecting more than 100K hits per day to the server, including 2
> additional queries (insert to counter, insert with remote_host info
> {get_env}) per hit.
> 
> I'm running this thing on the same net as a very qualified NT IIS admin
> who's been running several business sites for several years without
> interruptions (don't ask me how, I guess he's just very good w/ the M$
> stuff).
> 
> I really don't want this thing to fall on it's a$$ if/when it starts
> sucking in the customers.
> 
> If anyone can point me toward some obvious pitfalls on this, please offer
> any pointers you wish.  I'm feeling good about this, but, I just want to
> remove any glaring doubts.
> 
> The machine will run exclusively mysql/httpd/local telnet, so, I think I'm
> covering most of my bases.
> 
> Thanks in advance for any suggestions.  I've been working my a$$ off for
> this opportunity to promote Linux/UNIX/MySQL for 2 years, now, and, this
> is it.
> 
> Peace All,
> Van
>

Make sure to benchmark your configuration first. I do not think that
kernel tweaking will make that much of a difference, unless you start
running out of some system resources - if you do, just increase the
appropriate resources.

Focus your optimizations on good table design, efficient queries, and
properly configuring MySQL parameters.

-- 
Sasha Pachev
http://www.sashanet.com
Thread
DISTINCT keyword not doing as I expect...DC Mahoney15 Jul
  • Re: DISTINCT keyword not doing as I expect...Paul DuBois15 Jul
    • Re: DISTINCT keyword not doing as I expect...jkraai15 Jul
  • Optimizing a P133 for MySQLVan15 Jul
    • Re: Optimizing a P133 for MySQLSasha Pachev15 Jul
    • Re: Optimizing a P133 for MySQLJunster15 Jul
      • Re: Optimizing a P133 for MySQLOrlando Andico15 Jul
        • Re: Optimizing a P133 for MySQLJunster15 Jul
          • Re: Optimizing a P133 for MySQLOrlando Andico15 Jul
            • Re: Optimizing a P133 for MySQLEd Carp15 Jul
              • Re: Optimizing a P133 for MySQLOrlando Andico15 Jul
                • Re: Optimizing a P133 for MySQLJon Drukman15 Jul
                  • Re: Optimizing a P133 for MySQLOrlando Andico15 Jul
                • Re: Optimizing a P133 for MySQLChris Trown15 Jul
                  • Re: Optimizing a P133 for MySQLEd Carp15 Jul
                    • Re: Optimizing a P133 for MySQLOrlando Andico16 Jul
                      • Re: Optimizing a P133 for MySQLVan16 Jul
              • Re: Optimizing a P133 for MySQLVan15 Jul
              • Re: Optimizing a P133 for MySQLT├Ánu Samuel16 Jul
                • Re: Optimizing a P133 for MySQLJon Drukman16 Jul
        • Re: Optimizing a P133 for MySQLJon Drukman15 Jul
      • Searching across multiple tablesBrighten Godfrey15 Jul
        • Searching across multiple tablessinisa15 Jul
        • Re: Searching across multiple tablesChristian Mack21 Jul
    • Re: Optimizing a P133 for MySQL {Interim Results}Van16 Jul
      • Re: Optimizing a P133 for MySQL {Interim Results}Ed Carp16 Jul
  • DISTINCT keyword not doing as I expect...Michael Widenius30 Jul
Re: Optimizing a P133 for MySQLDerick H Siddoway16 Jul
Re: Optimizing a P133 for MySQLKen Farwell16 Jul