DSXchange

Use intelligent job control, read your Sequential file value and pass it to jobs as a parameter, use the parameter in your WHERE clause to avoid spooling unnecessary rows, and have your job control update your Sequential file when finished.

KPReddy wrote:Thanks Kenneth,
To use sqlldr to load data on remote DB load files need not be on DB server. I have invoked sqlldr from DS and loaded data on remote DB.

Only for conventional load does this work.
DIRECT path requires load files to be co-resident with the database instance.

Your issue is simple. The load files have to physically be on the same server as the database. Also, you must run sqlldr on that server, not invoked remotely. Therefore, to load your files you must ftp them to the database server and then rsh execute sqlldr. In order to use DIRECT path, you must und...

When you installed the new version of DataStage, did you use the Enterprise (PX) license or Server license? I've seen this issue when an Enterprise license was used but the Server license was left blank. You actually couldn't have Server jobs, just PX jobs and Sequencers (which are really Server job...

A lot of functionality is available within the dsjob command line program. You can query a project for a list of jobs, then cycle thru the list of jobs and request the current status of each.

There's no functionality inherent in the products for monitoring file availability. There are various methods, some using Unix scripts, others using DS BASIC or Sequencers, for detecting the presence of files. No matter what, you will program your own solution. As far as what I would do, I would wri...

Reasons inserting into a table progressively degrades, any number of the following: indexes, FK constraints, rollback, commit, triggers, tablespaces, partitions. Triggers are huge impactors of insert performance. Then again, what is your SQL? Insert only or are there inserts and updates? Take your t...

Methinks you are bumping up against some limit, such as a 32BIT hash file being stuffed too full, database rollback exceeded, snapshot too old, etc. Can you describe the nature of the job that blows up and paste all yellow and red error messages?

Have you twiddled with the metadata in the target stage column? It sounds like it's complaining about VARCHAR(315), maybe a LONG VARCHAR, TEXT, or something similar is required on DB2 CLI for text larger than 255.

Try to set the data type to VARCHAR and the width of the column to 4000. DS does not support CLOBS, so you'll have to fool the stage and see if it will bind using VARCHAR. If it does (I don't know your Oracle release or DS version) then you will be okay to pull upto 4000 chars from the column. Try u...

On Server jobs the data types in a Sequential file are irrelevant for reading the data. All that matters is column names, order, and size (if fixed width). Therefore, thru some simple Excel spreadsheet manipulation, you could simply take your column names and create a DDL statement using VARCHAR and...

Methinks you have a permissions issue. The userid that imports the job must belong to the same group that executes the job, provided your umask setting is 002. Since every job has supporting files that keep status, log, and configuration information, different users may not share access to the suppo...

Server or Parallel job, which stages (ODBC,CLI, PX?)

Since I don't know your job design, consider doing the following. Put an @FALSE constraint in the first transformer after the ODBC stage. This will cause the job to run and have all rows transferred from the source database to the transformer and then be thrown away. If the query completes successfu...

Turn it into DDL and create a table. All that really matters for a Sequential file stage is column name, position, length (if fixed width then it's really the DISPLAY column that's used, and scale. The data type can be all VARCHAR if it's a delimited file.