I found a parameter called TimeBetweenWatchDogCheck for a MySql Cluster. Could you give an example for not clustered MySql servers? Info: Our application is not a webbased one.
Regards, André -----Ursprüngliche Nachricht----- Von: Dan Nelson [mailto:[EMAIL PROTECTED] Gesendet: Freitag, 28. Juli 2006 19:00 An: André Ress Cc: mysql@lists.mysql.com Betreff: Re: Threads of terminated clientapplications In the last episode (Jul 28), André Ress said: > We have a problem with abortet clients and long time querys. > > If our application performs a long time select statement on the mySql > database and the application is abortet, the query still runs! We > tried several time-out variables like wait_timeout, > interactive_timeout or connection_timeout but without success. What > we need is, that the connection thread gets terminated after the > client application was abnormaly terminated. We currently use a > mySQL V5.0.20 server and have the problem under windows and linux os. > > Terminating the thread with KILL is not possible option, it has to be > terminated automatically. You might want to file an enhancement bugreport on this. It would probably require a watchdog mysqld thread that does a select() or poll() on all active client connections periodically (every 60 seconds or so?) to see whether they have been closed at the remote end, and if they have, it sends a kill to the thread currently running a query. Note that if your application is a web-based one, you will also have to do this check yourself (i.e. poll the browser socket so you know when they have hit the Stop button, and if they have, then close your connection to the database). -- Dan Nelson [EMAIL PROTECTED] -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
