Search found 42189 matches
- Thu May 31, 2007 6:31 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Change_capture vs CRC32
- Replies: 15
- Views: 11350
Welcome! :D Always best to start your own thread and then reference / link to any older thread that inspired you. That way we have a clue what your 'particulars' are - O/S, DS version, etc. So I'd propose you do that - start a new thread in the proper forum based on your job type, Server or PX, as t...
- Thu May 31, 2007 6:24 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Performance from 86 rows to 1 row per second
- Replies: 11
- Views: 4254
Re: Adding the hash file solved the problem
I added a hash file between the oracle lookup stage and the transformer. Always best to give as much information as possible from the beginning. You hadn't mentioned any Oracle lookup that I recall. So, you replaced an OCI lookup with a Hashed File lookup? Hashed files are the heart and soul of the...
- Thu May 31, 2007 6:20 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Reading parameter from csv file
- Replies: 5
- Views: 1581
If you are on a 7.5.x version and thus have the new Start Loop, End Loop and UserVariables stages you can build a looping Sequence job. You would first get a list of your parameters from the file using the UserVariables stage and then it would pass them, one at a time, to your Loop that runs the pro...
- Thu May 31, 2007 6:15 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: passing a parameter
- Replies: 11
- Views: 3571
Wolfgang, I think you've missed something. This is perfectly valid 'job control' code and seems to have been cloned from the 'Add Job' function. Ok, in pondering this - we've both made assumptions. I assumed a separate 'job control' job being run that in turn runs the other job. Luc, is that the ca...
- Thu May 31, 2007 6:11 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: passing a parameter
- Replies: 11
- Views: 3571
Wolfgang, I think you've missed something. This is perfectly valid 'job control' code and seems to have been cloned from the 'Add Job' function. One problem, Luc, is the fact that you don't check the success or failure of your operations before the job is run. You should check to see if the DSSetPar...
- Thu May 31, 2007 5:58 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: .txt File without delimiter
- Replies: 4
- Views: 1844
You won't be able to 'omit' the first two records, but you could easily constrain your job to only handle from record 3 onward. For example...
Code: Select all
@INROWNUM > 2- Thu May 31, 2007 5:52 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: getting today's date in datastage
- Replies: 7
- Views: 19607
Re: getting today's date in datastage
satyanarayana wrote:you can use DSJobStartTimestamp OR DSJobStartDate
Correct, those are macros that would give you a static value, if that's what one needed.
- Thu May 31, 2007 5:50 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: getting today's date in datastage
- Replies: 7
- Views: 19607
Re: getting today's date in datastage
How shall i get the system date in datastage .. . Is it an function or a datastage defined constant Yes. You've got both. Check the online help for: Date() @DATE Time() @TIME TimeDate() These are various ways to get the time and/or date from the system. What Vijaydev mention...
- Wed May 30, 2007 6:32 pm
- Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
- Topic: Time out error encountered while running a job sequence
- Replies: 4
- Views: 1471
I'd wager it's all about the timing. Too much going on when the jobs were first run, but after taking the time to compile them after the failure, they run fine because other processes have completed and they were able to start within the timeout period. :D I really doubt the recompile 'fixed' anythi...
- Wed May 30, 2007 6:28 pm
- Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
- Topic: Encryption
- Replies: 4
- Views: 911
The data is coming to you 'encrypted' or do you need to perform the encryption during the load? If you don't need to do the encryption or any kind of decryption during the processing, I don't see why it couldn't be treated like any other 'data' - simply stored and retrieved from the database. Would ...
- Wed May 30, 2007 6:23 pm
- Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
- Topic: Message Digest function in Datastage
- Replies: 5
- Views: 1791
You don't necessarily need any kind of hashing algorithm. We do something like this with a database table and a generated surrogate key. The 'string' value is the PK and DS jobs maintain the information in the database. New string values get assigned new surrogate ids and lookups get you the surroga...
- Wed May 30, 2007 3:00 pm
- Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
- Topic: Sequencer
- Replies: 5
- Views: 947
- Wed May 30, 2007 7:04 am
- Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
- Topic: Job Warning
- Replies: 3
- Views: 1463
- Wed May 30, 2007 7:01 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: wait for file stage
- Replies: 2
- Views: 997
- Wed May 30, 2007 6:36 am
- Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
- Topic: Performance from 86 rows to 1 row per second
- Replies: 11
- Views: 4254
Re: Performance from 86 rows to 1 row per second
When inserting en updating more and more data to a oracle table the performance slowly decreases from 86 rows to eventually a 3/1 rows per second. If "inserting and updating" means you are using one of the 'combo' Update actions, then that is probably the heart of your problem. Never mind the fact ...