Search found 53125 matches
- Wed Jun 08, 2005 2:47 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Test for empty hashed file
- Replies: 5
- Views: 1885
Put a UV stage on the end of a stream input link. The link has one column, an integer. Derive the column as COUNT(*) (or COUNT(@ID)) and specify the hashed file name as the table name. In the Transformer stage fed by this stream input link, detect whether the value of this column is 0 (in an output ...
- Wed Jun 08, 2005 2:44 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Distributed Hash Files
- Replies: 14
- Views: 4450
There is no DELETE.DF (and the I-descriptor does not need to be in the VOC file - where did you read that?). You could try reinstating the DF (verbs are VERIFY.DF and REBUILD.DF). One problem that you may have is that each of the part files has a block of data (the "partfile" block) indicating that ...
- Wed Jun 08, 2005 2:39 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: how to use Account Name in hash file
- Replies: 9
- Views: 3270
VOC is a hashed file, as stated, and can have 12 or more different record formats ranging from two fields to a theoretically infinite number. Most have between two and nine fields. That's really why you can't read it; the metadata to describe such a beast are horrendous! Why do you want to read the ...
- Wed Jun 08, 2005 2:32 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Stored Procedure in ODBC Stage
- Replies: 1
- Views: 767
You can not create a Stored Procedure in an ODBC stage. The ODBC stage can use an existing Stored Procedure to perform its activity, but that Stored Procedure must already exist in the database, must operate on a record set, and must have been imported into the DataStage Repository (so that the numb...
- Wed Jun 08, 2005 2:28 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: NOTFOUND as field name
- Replies: 4
- Views: 1026
Depending on the stage type at the other end of the link, you can effect a column name change by using a column name in DataStage (maybe FRED) and its derivation referring to the real column name (table.NOTFOUND). Or your derivation can explicitly specify an alias ( for example table.NOTFOUND AS FRE...
- Wed Jun 08, 2005 2:22 am
- Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
- Topic: Pivot Stage
- Replies: 7
- Views: 2526
- Wed Jun 08, 2005 2:20 am
- Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
- Topic: IsValid Function
- Replies: 1
- Views: 1487
Because twos-complement storage is used for integers, any negative integer that is small enough can be reported as a valid unsigned integer. For example the signed TinyInt -1 (int8) has the same bit pattern as the unsigned equivalent 128 (uint8), namely "10000000". You need the IsValid test and anot...
- Wed Jun 08, 2005 2:16 am
- Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
- Topic: Convert VarBinary Datatype to string
- Replies: 7
- Views: 6028
You have no control over the metadata in the rejects link of a Sequential File stage. You get one column; the raw data record (which does not match the metadata on the input link) and its type is VarBinary. If you want to change this you need to feed it into a downstream stage in which you can chang...
- Wed Jun 08, 2005 2:14 am
- Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
- Topic: Oracle 8 error 1002
- Replies: 1
- Views: 1014
- Wed Jun 08, 2005 2:09 am
- Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
- Topic: How to reterive Metadata informatrion
- Replies: 3
- Views: 1189
- Tue Jun 07, 2005 3:09 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Distributed Hash Files
- Replies: 14
- Views: 4450
SYSTEM requires a key of the form partno:delimiter:remainder. For example: 1/7603, 2/1311 and so on. The delimiter character is specified in the algorithm, for example SYSTEM '/' for the above keys. So, if the first column of your key can be guaranteed to contain the part numbers (1 through 10?), th...
- Tue Jun 07, 2005 3:04 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Update tabla through Oracle Stage
- Replies: 11
- Views: 3323
Within DataStage use different column names. The first column in the grid replaces the parameter marker :1, the second replaces :2, and so on. Make sure that the one to appear in the WHERE clause is marked as Key. Choose user-defined SQL as the preferred method, then create the UPDATE statement that...
- Tue Jun 07, 2005 2:59 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Error while selecting column form a hash file
- Replies: 4
- Views: 946
- Tue Jun 07, 2005 2:57 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Distributed Hash Files
- Replies: 14
- Views: 4450
If you have 10 hashed files (with identical record structure and some characteristic of the key that consistently identifies a part file) then you can construct a Distributed file. However, your partitioning algorithm only specifies a two part Distributed file. Which do you want; two or 10 parts? Im...
- Tue Jun 07, 2005 2:44 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Job completion status
- Replies: 5
- Views: 1454