I have a Job Design that will run successfully with small volume but as soon as I try larger volume it crashes with the below error.  This happens both on a windows machine with the job being run via TOS,  and also after deploying on a Unix box.

It looks like records (1 million) are being read ok from the input sources, but that maybe the tmap is unable to cope.   I have used the 'store on disk' option,  and have experimented with the 'Max Buffer Size, nb of rows.   I have also tried setting the Xmx param higher in both the TOS preferences,  and the Unix wrapper script.  Again without success.

Any suggestions appreciated.







Starting job Recon_Dn001schedule at 10:43 22/07/2010.

[statistics] connecting to socket on port 3572
[statistics] connected
Buffer marked at index (1-Lookup) 2000000, to avoid a heap space memory error try to increase the JVM Xmx parameter.
#
# A fatal error has been detected by the Java Runtime Environment:
#
# java.lang.OutOfMemoryError: requested 4092 bytes for char in C:\BUILD_AREA\jdk6_20\hotspot\src\share\vm\utilities\stack.inline.hpp. Out of swap space?
#
#  Internal Error (allocation.inline.hpp:39), pid=4240, tid=2812
#  Error: char in C:\BUILD_AREA\jdk6_20\hotspot\src\share\vm\utilities\stack.inline.hpp
#
# JRE version: 6.0_20-b02
# Java VM: Java HotSpot(TM) Client VM (16.3-b01 mixed mode windows-x86 )
# An error report file with more information is saved as:
# C:\Talend\TOS-Win32-r43696-V4.0.2\hs_err_pid4240.log
#
# If you would like to submit a bug report, please visit:
#   http://java.sun.com/webapps/bugreport/crash.jsp
#
Job Recon_Dn001schedule ended at 10:44 22/07/2010. [exit code=1]