DSXchange

My solution is the same as Vincent's. In addition, to make it really clear (graphically) what's happening in the (7.5) job sequence, add a Terminator to send "Stop" requests to any running jobs.

Preallocating the disk space when you create the empty hashed file will help. Use MINIMUM.MODULUS if creating a dynamic hashed file, or modulo if creating a static hashed file.

You can certainly achieve the result with two Transformer stages, passing the MinValue along the link between them. Or you could construct a user-defined nested SQL statement where the inner query returns the result used in the outer query. This is standard database practice and, used in this fashio...

Each tool is a good tool. Each tool is in Gartner's top right quadrant. They involve a different mind set during design (simplified as top-down versus bottom-up) but should end up with the same solution from the same plan. Any person may prefer one over the other. Here you're likely to get a bias to...

DataStage has a better splash screen, DataStage is easier to say while you are holding your tongue and DataStage rhymes with MetaStage, ProfileStage and QualityStage. I think you'll find Informatica rhymes with none of those products. Isn't there a Metabroker that allows MetaStage to get stuff from...

Yes, of course, particularly if you are depending on the partitioning being the same, for example to perform a lookup.

Whose documentation? Where? That's just crazy! The whole point of using partitioned data is to make use of partition parallelism. I think that you will find that any such assertion is subject to some kind of conditional clause.

You might need the -n option for sorting numeric fields with the sort command, otherwise they're sorted as strings.

SETFILE pathname VOCname OVERWRITING creates an entry called VOCname in the VOC file as a pointer to the hashed file whose pathname is specified by pathname - it is the name in the VOC file that you must use in the TRANS function. TRANS(VOCname, keyvalue, field#, code) opens the hashed file referre...

You will need to create a VOC pointer to a directory-pathed hashed file. Search the forum for SETFILE.

Let me stress again that there are no job log files (in the sense that they can be read using operating system utilities). DataStage maintains job log tables (which are implemented as hashed files, at least in current versions).

You DO have control over data - you're the ETL person!!! You can determine what to put in a hashed file - apply an Upcase() function as it goes in. You can determine how to lookup those values - use an Upcase() function in the Key Expression in the Transformer. None of this affects the external data.

It IS possible in a server job, however it's not possible with a Hashed File stage (where you must lookup the key, the whole key, and exactly the key). You can use a UV stage to refer to the hashed file and do SQL-style lookups, but this won't be fast. It can be improved by creating indexes on the h...

All the indexing in the Repository is hard coded with the old pathnames. Entries in the system tables (SQL Catalog) also probably haven't been fixed to point to the new locations. As others have said, you can't just move a project like that. Remedy the index problem with SET.INDEX commands (reindexi...

Given that everything's in the correct sequence, the Fmt() function with a mask seems to me to be an easy solution. The first example loses the fractional seconds, the second example keeps six digits' worth of them. Fmt(Inlink.TheCol, "L####-##-## ##:##:##") Fmt(Inlink.TheCol, ...