Maxscale - after update Prepared_stmt_count is getting very high

Hi,

After we upgraded our Maxscale instance from 6.3 to 23.08 we had issues because the max_prepared_stmt_count limit was reached. And we started to receive "Can't create more than max_prepared_stmt_count statements (current value: 16382)" errors.

The Com_stmt_close counter was going up very slow, but could not hold up.

A friend of mine has seen this behaviour with ProxySQL too; the cause here was that multiplexed connections do not close the connection after an prepared statement (insert in this case).

I downgraded the instance to 6.3 and it works again without problems (max_prepared_stmt_count is stable at 80)

I already checked changelogs for changes, but could not find any.

This is my config:

# Globals
[maxscale]
threads=auto
log_debug=1

# Admin is enabled but behind nginx proxy, so thats why only http
admin_secure_gui      = false

[db1]
type=server
address=172.31.32.21
port=3306
protocol=MySQLBackend

[db2]
type=server
address=172.31.32.22
port=3306
protocol=MySQLBackend

[db3]
type=server
address=172.31.32.23
port=3306
protocol=MySQLBackend

[Galera-Monitor]
type=monitor
module=galeramon
servers=db1,db2,db3
user=maxscale
password=xxxxxxx
monitor_interval=1000

# Galera router service
[Galera-Service]
type=service
router=readwritesplit
servers=db1,db2,db3
user=maxscale
password=xxxxxxx
enable_root_user=1
max_connections=1800
connection_timeout=60

# Galera cluster listener
[Galera-Listener]
type=listener
service=Galera-Service
protocol=MariaDBClient
port=3306
address=0.0.0.0