Page 1 of 1

problem of join stage

Posted: Wed May 07, 2008 9:05 pm
by yulangok
We developed a complex job with several lookup stages.Now we should update the job in view of the attribute of lookup and join stage, and we decide to use join stage to repalce lookup stage. As I am going to finish my job and I use the join stage to replace the last lookup stage, a problem appears.The error message is "cs}}},0: Failure during execution of operator logic. " and "cs}}},0: Fatal Error: Fork failed: Not enough space". My job is so complex that I think that the reason may be that it exceeds the capability of the datastage server.
My dear friends, please give me advice.
Many thanks!

Posted: Wed May 07, 2008 9:59 pm
by devidotcom
Why don't you seat calmly and check what kind of join types you have applied.

The Operator logic error appears when the code has some faults.... I mean coded wrongly.

I understand as the job is complex its a pain but thats the solution.

Re: problem of join stage

Posted: Thu May 08, 2008 4:09 am
by ArndW
yulangok wrote:Fatal Error: Fork failed: Not enough space
is a message from your OS (you can look that up using google). What OS do you have, what are the ulimit settings, how many nodes in your APT_CONFIG file?

Posted: Thu May 08, 2008 9:39 pm
by yulangok
MY OS is WinXp, and my sever OS is HP UNXI. There are 4 notes in my APT_CONFIG file. Please tell me how I check ulimit setteing.

Posted: Thu May 08, 2008 10:14 pm
by ray.wurlod
Please post something that explains how this problem was resolved.

Posted: Fri May 09, 2008 1:58 am
by ArndW
As a datastage user enter "ulimit -a" at UNIX. Better yet, add "ulimit -a" to a ExecSH call in your job's before-job routine call and view the result in the log file.