4.0.x and Replication on FreeBSD still fragile...

MySQL 4.0.x (various builds from BitKeeper over the last month or so)
still exhibits fragile behavior on FreeBSD.

How-To-Repeat: Setup a 4.0.x slave running on FreeBSD. Break
replication somehow so that the SQL thread stops. Then
try to fix it by doing:

SET SQL_SLAVE_SKIP_COUNTER = 1
SLAVE START SQL_THREAD

The second command often will never return. Sometimes
it does but takes a long time (2 minutes or more). But
most of the time it never returns. The result is that
there's no reliable way of fixing a 4.0.x slave once it
has encountered a problem. I just lost a machine to
this again. I was hoping it had been fixed, but I
guess not.

Similarly, you can stop the IO thread and then just try
running a "SLAVE START" command. It will start the IO
thread, but the SQL thread will never start and the
command will never return. However, I know the IO
thread runs because the relay file begins to grow.

If you need access to a FreeBSD machine for testing, let me know.
This problem has been present in every release and build of MySQL 4
that I've tried.

Thanks,

Jeremy
--
Jeremy D. Zawodny,
Technical Yahoo - Yahoo Finance
Desk: (408) 349-7878 Fax: (408) 349-5454 Cell: (408) 685-5936

------------------------------------------------------------ ---------
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 bugs-thread12129@lists.mysql.com
To unsubscribe, e-mail