Re: connector mxj can't init db on linux

Hello Andreas,

For whatever reason, I don't see the attachment. 

Could you please re-send it? 

On Thu, 2006-05-04 at 17:03 +0200, Andreas Schlicker wrote:
> Hi Eric,
> 
> 
> > I'd like to ask you to obtain some more data for me. On a machine which
> > is having the "InnoDB: Unable to lock ./ibdata1, error: 11" problem, I'd
> > like you to do the following, and capture all of the screen output for
> > each, and send it to us.
> > 
> > 1) Start by removing the existing basedir and datadir directories.
> > 
> > 2) Next run the first of your test programs. 
> > 
> > 3) do an "ls -l" of the datadir
> > 4) run the second program (we expect this to fail)
> > 
> > 5) do another "ls -l" of the datadir
> 
> 
> There is absolutely no differences between the two 'ls -l' runs. I also 
> did an 'ls -l data/* (attached) without seeing a difference.
> 

I am not surprised that there is no difference between the two "ls -l"
runs, but if there were a difference, that would be very interesting.

What I am hoping will be much more informative will be the screen output
of Connector/MXJ (and mysqld) from the two test programs.

Perhaps we shall see some message about available memory. Whatever the
problem is on these machines, I'm hoping that something in the screen
output will be informative. 

The "InnoDB: Unable to lock ./ibdata1, error: 11" suggests that there is
some resource availability problem related to innodb trying to lock
ibdata1. Error 11 is "EAGAIN", which is "Error try AGAIN".

perhaps the attempt to lock ibdata1 executes a a non-blocking read
instruction when there is (at that time) nothing for it to read. Or
maybe this is being caused because the OS is unable to satisfy our
system call, due to a lack of resources at that moment -- for example,
we may have run out of memory. Or perhaps it is something else entirely.

I'm hoping the "ls -l" will at least tell us something about the
"ibdata1" file. Alternatively, if it is memory related, then we might
have a guess as to why we see the problem with the JVM running but not
when you executed the "launches" and "kills" of mysqld by hand. However,
I really do not know. 

Still, I remain hopeful that something in the screen output from a
"fresh" run will give us some good insight.

Cheers,
 -- Eric

> Regards,
> Andreas
> 
> -- 
> MySQL Java Mailing List
> For list archives: http://lists.mysql.com/java
> To unsubscribe:    http://lists.mysql.com/[email protected]
-- 
Eric Herman, Software Developer
MySQL AB, www.mysql.com
Mobile: +1 206 913 8942

Are you MySQL certified?  www.mysql.com/certification


-- 
MySQL Java Mailing List
For list archives: http://lists.mysql.com/java
To unsubscribe:    http://lists.mysql.com/[email protected]