MySQL : Problème mémoire serveur Linux

Bonjour,

J'ai un problème avec un process MySQL qui utilise de plus en plus de mémoire et qui ne la libère pas...
Au bout d'un moment il utilise toute la mémoire disponible sur le serveur...


Version MySQL : 5.1.51
OS : Rhel 5

Voilà ce que me retourne la commande "ps aux" :

xxxxx 12496 4.3 38.9 10832240 9609528 ? Sl Aug05 1448:36 /xxxxx/xxxx/xxxxxx/xxxxxxx/mysqld --defaults-file=/xxxx/xxxx/xxxxx/xxxxxx/my.cnf --basedir=/xxxx/xxxxx/xxxx --datadir=/xxxx/xxxxx/xxxx --user=xxxxx --ssl --log-error=/xxxx/xxxxx/xxxx/error.log --pid-file=/xxxx/xxxxx/xxxx/mysql.pid --socket=/xxxx/xxxxx/xxxx/mysql.sock --port=xxxx

Soit 10Go de Mémoire !
Le problème c'est que hormis que c'est énorme, cela augmente régulièrement jusqu'à saturation complète du système...


Voici mon fichier de configuration my.cnf :

[client]
port = xxxx

socket = /xxxx/xxxxxx/mysql.sock

[mysqld]
port = xxxx
socket = /xxxx/xxxxxx/mysql.sock
pid-file = /xxxx/xxxxxx/mysql.pid
datadir = /xxxx/xxxxxx
tmpdir = /xxxx/xxxxxx/tmp

user = xxxxxx

event_scheduler = ON

character-set-server = latin1
collation-server = latin1_swedish_ci

skip-external-locking


key_buffer_size = 256M

max_allowed_packet = 1M

table_open_cache = 256

sort_buffer_size = 1M

read_buffer_size = 1M

read_rnd_buffer_size = 4M

myisam_sort_buffer_size = 64M

thread_cache_size = 8

query_cache_size = 16M

secure-auth
secure-file-priv = /xxxx/xxxxxx

max_connections = 104

max_user_connections = 50

ssl_ca = /xxxx/xxxxxx
ssl_cert = /xxxx/xxxxxx
ssl_key = /xxxx/xxxxxx

server-id = 1
#relay-log = /xxxx/xxxxxx/mysql-relay-bin
#relay-log-index = /xxxx/xxxxxx/mysql-relay-bin.index
#master-info-file = /xxxx/xxxxxx/master.info
#relay-log-info-file = /xxxx/xxxxxx/relay-log.info
#log-slave-updates
#replicate-same-server-id=0
#auto_increment_increment=2
#auto_increment_offset=1


general_log = OFF
general_log_file = /xxxx/xxxxxx/log-general.log

log-bin = /xxxx/xxxxxx/log/bin
binlog_format = mixed
log-error = /xxxx/xxxxxx/log/error.log
slow_query_log_file = /xxxx/xxxxxx/log/slow-queries.log

long_query_time = 4

expire_logs_days = 7


innodb_file_per_table

innodb_data_home_dir = /xxxx/xxxxxx/data/

innodb_data_file_path = ibdata1:10M:autoextend

innodb_log_group_home_dir = /xxxx/xxxxxx/log/

innodb_buffer_pool_size = 256M

innodb_additional_mem_pool_size = 20M

innodb_log_file_size = 100M

innodb_log_buffer_size = 8M

innodb_flush_log_at_trx_commit = 1

innodb_lock_wait_timeout = 50

[mysqldump]
quick
max_allowed_packet = 16M

[mysql]
no-auto-rehash
#safe-updates

[myisamchk]
key_buffer_size = 128M

sort_buffer_size = 128M

read_buffer = 2M
write_buffer = 2M

[mysqlhotcopy]
interactive-timeout





Le moteur de stockage de toutes les tables est MyISAM.

Avez-vous une idée de la raison d'une telle consommation de mémoire ?
Ne peut-on pas fixer un "quota" de concommation de la mémoire (ceci afin d'éviter de saturer le serveur) ?


D'avance merci