Hi all,
I begin to validate MySQL with InnoDB on our platform. I use a DBI perl
test which manages a data pool reflecting exactly the rows stored in DB.
This test internally builds some new data to insert and chooses rows to
delete and update (the tables are chosen randomly). Then all these
operations are executed in one transaction. If it fails, a rollback is
executed in DB and in client memory. Client also check that all data in
its pool are in the DB. That's my test...
It works very well with one client, but with two... things become
harder! clients received many 'Table handler error 1000000' and slow
down terrifically.
I try the same test with 3 clients on MyISAM tables and there is no
noticeable problems.
I have reduced number of queries per transaction. I have changed
innodb_unix_file_flush_method to O_DSYNC. I have increased buffer pool
size and log buffer size but I noticed onlky small improvements. I'm
sure I forgot something, but I don't see what ! does somebody know ??
Denis
--
Denis Pithon phone +33 (0) 1 41 40 02 13
Software Engineer fax +33 (0) 1 41 40 02 01
Lineo High Availability Group mail [EMAIL PROTECTED]
---------------------------------------------------------------------
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
