Hey folks,

I've looked around, but haven't seen anything similar to _quite_ the problem
I have.  The symptoms are thus:  memory usage for each mysqld process in top
creeps up to around 13%, 274MB, and stays there, no matter what I set my
config to.

79 processes: 76 sleeping, 3 running, 0 zombie, 0 stopped
CPU0 states: 100.0% user,  0.1% system,  0.0% nice,  0.0% idle
CPU1 states: 98.2% user,  1.2% system,  0.0% nice,  0.0% idle
CPU2 states:  1.1% user,  2.0% system,  0.0% nice, 96.2% idle
CPU3 states:  0.0% user,  0.1% system,  0.0% nice, 99.4% idle
Mem:  2064836K av, 2053068K used,   11768K free,       0K shrd,   15836K
buff
Swap: 2096440K av,   35312K used, 2061128K free                 1429524K
cached

  PID USER     PRI  NI  SIZE  RSS SHARE STAT %CPU %MEM   TIME COMMAND
18339 hunt      25   0  267M 267M   696 R    99.9 13.2  64:53 framework
  985 mysql     15   0  274M 259M 14460 S     0.0 12.8   0:00 mysqld
 1015 mysql     15   0  274M 259M 14460 S     0.0 12.8   0:00 mysqld
 1016 mysql     18   0  274M 259M 14460 S     0.0 12.8   0:00 mysqld
 1017 mysql     16   0  274M 259M 14460 S     0.0 12.8   0:00 mysqld
 1018 mysql     17   0  274M 259M 14460 S     0.0 12.8   0:00 mysqld
 1019 mysql     20   0  274M 259M 14460 S     0.0 12.8   0:00 mysqld
 1044 mysql     15   0  274M 259M 14460 S     0.0 12.8   0:00 mysqld
 1045 mysql     15   0  274M 259M 14460 S     0.0 12.8   0:09 mysqld
 1046 mysql     15   0  274M 259M 14460 S     0.0 12.8   0:00 mysqld
 1047 mysql     15   0  274M 259M 14460 S     0.0 12.8   0:09 mysqld
 1196 mysql     15   0  274M 259M 14460 S     0.0 12.8  16:14 mysqld
 1288 mysql     25   0  274M 259M 14460 S     0.0 12.8  29:06 mysqld
25425 mysql     25   0  274M 259M 14460 R    99.9 12.8  26:10 mysqld
29645 mysql     15   0  274M 259M 14460 S     0.0 12.8   1:04 mysqld

Now, I'm assuming some of that memory must be shared across
processes(threads?) since 14 processes * 12.8% is more than 100%, and
technically I still have 11.7M free.

My config has this:

>>>>>>>>>>>
set-variable = key_buffer=256M
set-variable = max_allowed_packet=1M
set-variable = table_cache=64
set-variable = sort_buffer=1M
set-variable = record_buffer=1M
set-variable = myisam_sort_buffer_size=64M
set-variable = thread_cache=8
# Try number of CPU's*2 for thread_concurrency
set-variable = thread_concurrency=8

>>>>>>>>>>>


Now, I've tried dropping all of the variables, cutting them in half, then
again, to no avail -- when I run a large query (via the DBI interface in
perl) the Mem usage jumps up to 13% (for all mysqld) and never comes back
down.  Most notibly I've dropped the key_buffer to 64M and the table_cache
to 32.

My thought was the query cache, since I thought that I read on the main site
that it will use about 13% per thread.  But no, I have that disabled:

| query_cache_limit               | 1048576
| query_cache_size                | 0
| query_cache_type                | ON

Size 0 supposedly means it is disabled -- or is this a bug?  Is is
query_cache_size 0' makes it think it is off but 'query_cache_type ON' makes
it cache anyway?  Regardless, flushing the tables and cache doesn't seem to
do anything.

I'm going to reduce the number of concurrent threads (it is set to 8 now)
but that's not the fix I want -- any helpful hints?

Actually, since it's set to 8, why do I get 14 processes?  Two of them are
parents to the other 12, but still?

Thanks,

Nick Elliott

Using MySQL 4.0.4-beta-log
Linux Kernel 2.4.18-17.7.xsmp


---------------------------------------------------------------------
Before posting, please check:
   http://www.mysql.com/manual.php   (the manual)
   http://lists.mysql.com/           (the list archive)

To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php

Reply via email to