DSXchange

No.

The key is already defined in the Lookup File Set. You don't need to do anything more.

The only time you need to define a key expression in the Lookup stage is when you're doing a lookup against a Data Set.

#$DBSchema#.TableName

The surrounding "#" characters are the way in DataStage that you specify a reference to a job parameter.

Naturally you have to have added $DBSchema to the job as a parameter.

But someone else on the early part of the learning curve might like to benefit from your experience. Can you advise what stage type you used and what expression(s) you used to solve this?

In parallel jobs you can declare the value to be hex. Someone (Vincent?) posted this undocumented trick some months back.

Row count.

The sequential file is being read on node number 0 but, since it's being processed in sequential mode, the node identifier doesn't need to be reported during progress. The total row count does include the node identifier, since it is likely to be the source for partitioning.

Because somewhere - almost certainly in a Transformer stage - you are trying to perform arithmetic or a numeric function with data that are not numeric.

You can view the code in file RT154/JOB.1062133866.DT.1384922442.TRANS2 on your server.

The two stages I mentioned allow you to specify some or all non-key columns as additional change detection columns.

A

There are quite a few differences. Your support provider will be able to supply a readme file if you haven't got 7.5.1A yet. Or you can get a "what's new" from Asecntial Developer Net

No. The SAP R/3 PACK is used to communicate with SAP R/3 (the SAP application server). The Web Services PACK contains stages for working with web-applications, for example the ability to work with web server logs. There are different kinds of Message Adapter, for example there is one for working wit...

2GB is a shitload (that's a technical term) of warning messages!!!

Your design won't be much different in PX. Probably a good thing to do is to select from the history table constrained on a time period, and ordered by any column in which you want to detect change. Then an approach using any of the change detection stage types, or even stage variables in a Transfor...

You should purge your log more frequently. Set auto purge. Don't let it grow past 2GB. If it's going to anyway, resize it to use 64-bit pointers.

The exact problem is that, when the error occurred, you already had as many dynamic hashed files open as permitted by the T30FILES tunable. You need to increase T30FILES to prevent recurrence of this problem. Remember that most of the Repository tables are dynamic hashed files; these are included in...

Static or dynamic - doesn't make any difference for the total volume of data to be stored. Choose dynamic for ease of maintenance. Choose static for optimum performance IFF (if and only if) the volume of data is precisely known and unchanging.

It must be the name of the hashed file in a Trans() function, not its pathname. You need to create a VOC pointer using SETFILE before the Trans() function will work. You would have determined this had you read the entry for Trans() in the manual.