Am I missing a configuration setting somewhere?
I have been working on a MySQL Max (3.23.52-max-nt) database running on W2K
Server.
I am using a MySQL client remotely over a VPN but unfortunately the client's
internet connection is severely overloaded and the connection dies on
occasion.
When this occurs the server process responsible for servicing the connection
lingers around forever and if this occurs during an update there are table
locks that also linger forever.
I have set the following variables in "my.cnf" and restarted the server to
have the server auto kill processes/connections after 60 seconds of
inactivity. I have confirmed the variable settings after the server restart
with SHOW VARIABLES
wait_timeout=60 (seconds)
interactive_timeout=60 (seconds)
BUT
the server never KILLS the processes! Even when the SHOW PROCESSLIST
command indicates that the process has been around for 3000+ seconds. Way
longer than the timeout values.
Am I missing some variable or command line setting to enable the server to
auto KILL processes?
**************
Reproduce as follows...
1) Configure server for short timeouts in my.cnf and restart server
set-variable = wait_timeout=60
set-variable = interactive_timeout=60
2) Start and connect a MySQL client to the server OVER A NETWORK LINK
3) Run the STATUS command to find your connection/process id so you know
which process to watch in step 8
4) Disconnect your network cable
5) Stop the MySQL client (so the server can't see the client shutdown, same
as if VPN hangs)
6) Reconnect the network cable
7) Start and connect a new MySQL client to the server
8) Run SHOW PROCESSLIST and watch the entry for your previous process id, it
never goes away! and its time just keeps incrementing.
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
