Variable PHP-mySQL response time on Suse 8.2 dual-processor

[Alan Hewat]

I have a relatively modest mySQL database with a PHP interface (70,000
entries). On a standard Dell Dimension 4400 (1.9 GHz with 512 Mbytes)
running Suse 8.1 a typical query takes 0.66 sec with no other system
load (http://icsd.ill.fr/)

On a new IBM dual processor XEON 2.4GHZ (1 GO RAM & 2 SCSI disks)
running Suse 8.2 the same query (with no other system load) takes
anywhere between 0.5 and 3.0 seconds - randomly variable response for
no obvious reason ! With only one query, only one processor is being
used, but why is the response so variable ? What can I try to track
down the problem ? After searching Google, I tried playing with renice
and my.cnf variables (below) but that made no difference.

skip-locking
set-variable = key_buffer=384M
set-variable = max_allowed_packet=1M
set-variable = table_cache=512
set-variable = sort_buffer=2M
set-variable = record_buffer=2M
set-variable = thread_cache=8
# Try number of CPU's*2 for thread_concurrency
set-variable = thread_concurrency=4
set-variable = myisam_sort_buffer_size=64M

I am not a database/system expert - just a scientist trying to use
these things.

Alan <He***@ill.fr>

[D. Pickering]

1. Do desc table name and make sure keys are the same and that there are
keys defined and used in the query
2. maybe try "explain yourentirequery" to see if it's using the defined keys
as expected
3. open a second shell and run w or top while doing the query....what for
other processes eating up the machine.
4. if needed, put the query into a loop to keep it running.
that's sure a disappointment isn't it. big machine and slower, and almost
worse: inconsistent.
dp

"Alan Hewat" <he***@ill.fr> wrote in message
news:9f**************************@posting.google.c om...

I have a relatively modest mySQL database with a PHP interface (70,000
entries). On a standard Dell Dimension 4400 (1.9 GHz with 512 Mbytes)
running Suse 8.1 a typical query takes 0.66 sec with no other system
load (http://icsd.ill.fr/)

On a new IBM dual processor XEON 2.4GHZ (1 GO RAM & 2 SCSI disks)
running Suse 8.2 the same query (with no other system load) takes
anywhere between 0.5 and 3.0 seconds - randomly variable response for
no obvious reason ! With only one query, only one processor is being
used, but why is the response so variable ? What can I try to track
down the problem ? After searching Google, I tried playing with renice
and my.cnf variables (below) but that made no difference.

skip-locking
set-variable = key_buffer=384M
set-variable = max_allowed_packet=1M
set-variable = table_cache=512
set-variable = sort_buffer=2M
set-variable = record_buffer=2M
set-variable = thread_cache=8
# Try number of CPU's*2 for thread_concurrency
set-variable = thread_concurrency=4
set-variable = myisam_sort_buffer_size=64M

I am not a database/system expert - just a scientist trying to use
these things.

Alan <He***@ill.fr>

[D. Pickering]

1. Do desc table name and make sure keys are the same and that there are
keys defined and used in the query
2. maybe try "explain yourentirequery" to see if it's using the defined keys
as expected
3. open a second shell and run w or top while doing the query....what for
other processes eating up the machine.
4. if needed, put the query into a loop to keep it running.
that's sure a disappointment isn't it. big machine and slower, and almost
worse: inconsistent.
dp

"Alan Hewat" <he***@ill.fr> wrote in message
news:9f**************************@posting.google.c om...

I have a relatively modest mySQL database with a PHP interface (70,000
entries). On a standard Dell Dimension 4400 (1.9 GHz with 512 Mbytes)
running Suse 8.1 a typical query takes 0.66 sec with no other system
load (http://icsd.ill.fr/)

On a new IBM dual processor XEON 2.4GHZ (1 GO RAM & 2 SCSI disks)
running Suse 8.2 the same query (with no other system load) takes
anywhere between 0.5 and 3.0 seconds - randomly variable response for
no obvious reason ! With only one query, only one processor is being
used, but why is the response so variable ? What can I try to track
down the problem ? After searching Google, I tried playing with renice
and my.cnf variables (below) but that made no difference.

skip-locking
set-variable = key_buffer=384M
set-variable = max_allowed_packet=1M
set-variable = table_cache=512
set-variable = sort_buffer=2M
set-variable = record_buffer=2M
set-variable = thread_cache=8
# Try number of CPU's*2 for thread_concurrency
set-variable = thread_concurrency=4
set-variable = myisam_sort_buffer_size=64M

I am not a database/system expert - just a scientist trying to use
these things.

Alan <He***@ill.fr>