DSXchange

I am trying to do left outer join on two files using merge stage...
I am getting the following error...
"Function 'construct_hash_table' failed
Unbalanced or unescaped quote character"..Any clue what I am missing here..?

Thanks

I have used the Merge stage on several projects to perform Left, Right, and Inner joins on sets of data.
The stage uses named files as sources, dynamically creates hashes under the covers to do the merge/joins, and then outputs data streams based on your keys and join definitions.
In my experience, the Merge designer interface (in 7.0 and 7.1) is somewhat buggy and difficult. Sometimes when you make changes to the stage, they do not save. Metadata manipulation for output streams is inconvenient and time consuming. There are hidden 'gotcha' issues that only appear after you build the job and try to compile, or worse, run it.

On the Stage tab, set Tracing Level to 1, that will write log file entries that may help you to track down your error.

Consider implementing your join(s) with regular hashes and Transformer stages, there's another thread going on right now on that subject.

Carter

Hi All,

This is the first time i tried using a Merge Stage.. In this job i was trying to merge two .csv files, employee data and Dept data files.. I used a complete Join as the condition in the Merge Stage.. The job got aborted with the following message in the director...

DataStage Job 721 Phantom 616
Program "DSP.Open": Line 122, Exception raised in GCI subroutine:
Access violation.
Attempting to Cleanup after ABORT raised in stage MergeStage..EmpDeptMerge
DataStage Phantom Aborting with @ABORT.CODE = 3


I couldn't get much information from the log.. has anyone tried using the merge stage before.. i am using Datastage 6.0 running in Windows NT box.. please help me in solving this issue.. is it advisible to use Merge stage.. or use hashfiles for joining of files.. please advice ...


Thanks
Sudharsanan

Access Violation means an attempt to access a memory location that doesn't belong to the application. This can occur, for example, by trying to put a character string N characters long into a buffer smaller than N characters (e.g. you've used a Char data type). It can also occur when there are unexpected NULLs in the data.

Thanks Ray for the input.. I tried looking into the job..

My Job design looks like this

Merge Stage -----> Tranformer Stage ---> Seqential file


I found that there are no problem with the data that i am using, if i use a sequential file in place of merge stage i was able to read the data.. also i don't have CHAR field in my definition.. also View data option in the merge stage is not avaliable for data verification...please let me know whether any issues with the merge stage...

I've never had any, but I don't use it much. Contact your support provider, or visit the eServices web site. Or search this forum.

Welcome aboard! :D

Thank you for posting a solution, and congratulations on the fact that your first post is a solution!