DSXchange

All the data in a single record are stored in a delimited string. What you're seeing in the "D" type entries is that, for example, the AMORT_RESIDUAL field is the 103rd delimited substring, and that there is one value of AMORT_RESIDUAL per record (the "S" in the "SM" place), and this is stored with ...

lebos wrote:Ray,

Don't you mean prior to version 5.2 for /.uvhome? I am using 5.2 and it makes use of /.dshome, not /.uvhome. According to the install manual 5.2 is where the change was made.

Larry

I stand corrected. It's been so long!!!

Not sure what you're asking here.

Ascential negotiated with IBM and obtained access to the partitioning algorithm for DB2, so that parallel jobs are able to specify "DB2" as the partitioning algorithm when extracting or loading data from/to DB2.

Is this what you were asking about?

Axcel PACK was the initial implementation of multiple instance server jobs. It was a separately charged part of the product at 5.1. It vanished (became free, with no need for separate licensing) at 5.2. Parallel Extender is Ascential's name for the Orchestrate technology they obtained by acquiring T...

Full table scan--->Seq files/Hash files---->Transformations--->Bulk Load . Don't forget that partial table scan may be even better. Only SELECT the rows and columns that you need to process. Restricting the rows will be assisted by indexes (of course), restricting the columns is simply only specify...

With versions 5.x and 7.x it's more of a case that you can get away with it, because the engines are different. Version 5.x is based on the UniVerse engine, installed in the directory pointed to by /.uvhome and, as you noted, uses port number 31438 for the RPC mechanism, and has shared memory identi...

I would report that as a bug. DataStage should not impose any restriction that is not imposed by the database server or its client software.

Alternatively, convert the dimensioned array to a dynamic array and use FINDSTR statement (which performs a partial match search). For example: Ans = @FALSE DA = "" * Convert dimensioned array to dynamic array MATPARSE DA FROM EQMASKTbl, @FM SETTING EQMASKEntries * Perform partial match se...

[quote="larryoceanview"]It probably would work to match on the string but it wouldn't handle the wildcard issue! Of course it would! That's what MATCH is for. Include wildcard characters in the match pattern (note that these are not Windows or SQL wildcard characters). In pattern matching ...

You might like to investigate the start command in DOS. This command starts your command in a separate window, and has options about priority and so on. Type help start | more at a DOS prompt. This may provide a mechanism to get around the issue of hanging NTVDM processes.

Well, that's not out of the question, either. Though I'm currently booked out until Easter, I expect it will take about three months to bed down (do the selection process) for a process of this magnitude. Oh, and to budget for my services of course.

Yes, you can have two (or more) links to the same table from DataStage. You may need to be careful with transactions locking the table, however. This is another reason for preferring a hashed file. Millions of key values is still not that large a hashed file. You only need the primary key values, be...

There are some simple test jobs you can create to determine the maximum capacity of your system. Job 1 tests read performance from a sequential file. This is the fastest possible job. Anything else will run more slowly . The Transformer stage has an output contraint of @FALSE, so the job discards ev...

If we were to use this methodology, wouldn't we still need to pull all of the unidata data from our database and then use a reference to find the changed/new records? No, if the logic for determining the changed records can be determined within UniData, you can simply add a selection criterion to y...

Nice thought, but you can't create F pointers from DataStage to UniData.