We have this same problem.

The JDK, even 1.4.2_02, still has problems with the new threading model in RH 
9 (NPTL).  This threading library will be standard in the 2.6 kernel.  The 
funny part is that Red Hat is touting the fact that the new threading speeds 
up Java programs a lot.  Seems that they all need to keep working on it all.

The workaround is to set an environment variable in the shell that is kicking 
off ant.

  'export LD_ASSUME_KERNEL = 2.4.1'

Your build should give you no trouble after that.


Eric Weidner
Out-of-the-Box
http://www.ejbsolutions.com

On Monday 24 November 2003 02:11 pm, Clifton Craig wrote:
> Hell all,
>
> I just don't know where else to turn on this problem. I have a rather
> involved and lengthy build process that has been running without fail until
> recently. I've been developing the build on WinXP and running it on both
> WinXP and Redhat 9 without fail. We have now setup a new development
> machine which I believe is Redhat 9 as well and I every time I run the
> build it pauses. I'm new to Linux so I'm not sure what state the process is
> in when it pauses. When I say it pauses I mean it does just that. It stops
> running and stops logging. I've tried many renditions of executing the
> build including running it in the background, running it in the foreground,
> with/without logging, logging to both the screen and disk, any and all
> combinations of the above. Normally what happens in all cases is it gets
> 1/2 way through and just stops in the middle of the EJB task for JOnAS. You
> see we have about 50 or so EJBs that are compiled/jarred for JOnAS v3.3 in
> this build using their current ejbjar task. It appears to stop at random
> points all the time. For example, if the beans were numbered 1-50 sometimes
> it will stop at bean 32, sometimes it will stop at bean 43, etc. It doesn't
> continually stop at the same bean. What's more if I re-run it after it
> stops it won't finish all the way. It only builds a few more beans and
> hangs up again. I have to run the build 3-4 times to get all the way
> through. I've tried using the "top" command when it pauses to see where the
> process is and it looks like its in a "SLEEP" status according to the man
> pages for top. Forgive me, as I don't know enough about Linux to understand
> why this is happening or even if the process is still alive. The same build
> runs on another (beefier) Linux box without fail. Has anyone ever run into
> such a problem before? Can anyone help me?
>
> Clifton C. Craig
> Intelligent Computer Systems
> A division of Global Beverage Group
> [EMAIL PROTECTED]
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to