DSXchange

There is no problem in doing this. You call up your job sequence with a "dsjob" command using a list of parameters, then do it again with different parameters for the second job.

There is no problem in doing this. You call up your job sequence with a "dsjob" command using a list of parameters, then do it again with different parameters for the second job.

I am not sure if you are aware of the performance impact of calling a perl script once per row in your job. Are you certain that you cannot emulate your regular expressions within DataStage?

If you declare it as VarChar(4000) in your job, what is the error message you get?

The job stays and waits until the .bat file returns control to it; so most likely the script is hanging (quite often the cause is an "input" that is required that never gets answered)

You cannot FTP a remote Excel file from inside DataStage and be able to parse the contents. Do you have an ODBC or other driver on your Linux server which can read excel files?

During that 30 minute wait time you should have a DBA monitor your session and find out what is happening.

What data type is "REC_TYPE" - I would wager it is "VarChar" - which isn't a fixed length.

Is it just some jobs or all jobs that cannot be changed and saved? With V8 the repository has become more complex and your error seems to indicate problems in the repository. One could try a complete shutdown and restart of the IIS application to see if that might make a difference.

A quick Google showed: Error code: 6091 TYPE_MISMATCH_BETWEEN_ATTRIBUTE_AND_ CONSTANT_ ON_EXPRESSION Cause: The type of the constant used for comparison in the expression does not match the type of the attribute. What is the column data type that you are lookup up on? Could there be a type mismatch ...

The sequential file stage can read and write fixed width files with or without delimiters. Just specify record length of "fixed" type file and delete the delimiter property.

Do you mean a C++ routine or a DS Basic routine? You posted in General, but as a PX type and the answer depends upon where you wish to execute a shell command from.

Please give a bit more information, as just the SQL error message doesn't help much. Does the error only occur in one job, or on one particular key? Does ODBC read for that table work (i.e. as a source, not a lookup)?

Apart from running in one node, can you try using "APT_ORA_WRITE_FILES" to keep the oracle loader files and to see if that can be run outside of DataStage without a problem?

will return a 0 if "BANK" is not in the string, otherwise the location of the first occurence of "BANK" in the string.