Database replication process stopped in server

UNIX HP-UX
1

Hi,

A database (Oracle) replication process was executed by the DBA team in one server (serverX). However, this replication process gets terminated, and there are no errors in the replication log. But there is error in the OS log files (syslog.log) :

Aug  8 16:51:47 L28dre02 sshd[10388]: subsystem request for sftp 
Aug  8 17:11:15 L28dre02 vmunix:
Aug  8 17:11:15 L28dre02 vmunix: Pid 11582 was killed due to failure in 
writing the signal context- possible stack overflow.
Aug  8 18:45:31 L28dre02 sshd[17209]: Accepted keyboard-interactive/pam 
for oracle from 192.168.29.23 port 50190 ssh2

From the error "stack overflow", I found that the solution was to edit the value of the parameter 'PTHREAD_DEFAULT_STACK_SIZE'.
However, I am not sure how/where to locate the file in the server.

Another article I read says that change that parameter value in the file :
$SYBASE/RAX-15_5/bin/ra.sh

However, when I run echo $SYBASE in the server to look for the path, the error is :

L28dre02:/home/oracle:INTDB> echo $SYBASE
sh: SYBASE: Parameter not set.
L28dre02:/home/oracle:INTDB>

How do I locate this parameter and change the value?

***The replication application used is 'DBVisit'.

2

Setup the profile of the user using ulimit -s to a higher value.
Possibly one suggested in your database vendor documentation.

If ulimit fails to set to <value>, you will have to adjust the maxssize kernel parameter, then use ulimit.

Be sure to read the manual pages of ulimit and maxssize parameter (this is what i like about hpux, man <kernel parameter> is golden;) )

Please for future note specify your shell (shell of the user the database processes runs under) and exact hpux version.
Folks here can be more exact in helping you with that information upfront.

I have not used the variable mentioned here so i cannot comment on that, give it a shot with standard tools mentioned above.

Hope that helps
Regards
Peasant.