Search found 55 matches

by prasannak
Fri Oct 31, 2008 12:24 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: duplicates using inrownum
Replies: 1
Views: 1379

duplicates using inrownum

We have an input dataset that feeds into a parallel transformer stage with as many output links as dimension lookups required. If 3 such lookups needs to be performed (and the matching surrogate key returned from the dimension), then the transformer output has 3 links. Within the transformer stage, ...
by prasannak
Thu Aug 28, 2008 5:10 pm
Forum: General
Topic: datatype mismatch warnings
Replies: 2
Views: 2115

Thanks ..Will try this out..
Is it similar to the cast function we have in oracle ?

Btw, I like the signature you have there :)
by prasannak
Thu Aug 28, 2008 4:04 pm
Forum: General
Topic: datatype mismatch warnings
Replies: 2
Views: 2115

datatype mismatch warnings

We have a oracle source number field coming in with number(19,4). In the target, the same has been defined as number(21,6) And there are a lot of number fields like this in our tables. Our ETL job expects a volume of 1 million rows everyday to be loaded from the source to the target. Now, I know tha...
by prasannak
Fri Aug 08, 2008 12:18 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Oracle Deadlocks
Replies: 3
Views: 3500

Hi Gaby, Apparently, load option using parallel stage has some bugs in datastage that has prevented us from using it... We have given this information to IBM and it is being ackonwledged by them it is indeed a bug and last heard from them was that they are working on it... No other oracle stage is r...
by prasannak
Fri Aug 08, 2008 12:12 pm
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Best approach for determining changes
Replies: 6
Views: 4170

CDC

Hi, We are facing similar CDC related performance issues in some of our heavy volume jobs... The scenario for us is : target dimension has about 20 million records. Incoming source data(oracle) has a total of 20 million records.. Since the source is not a fact, there seems to be no way to filter the...
by prasannak
Fri Aug 08, 2008 11:55 am
Forum: IBM<sup>®</sup> DataStage Enterprise Edition (Formerly Parallel Extender/PX)
Topic: Oracle Deadlocks
Replies: 3
Views: 3500

Oracle Deadlocks

Issue:- Oracle deadlock issue problem while inserting data into ORACLE DB. sqlcode is: -60 esql complaint: ORA-00060: deadlock detected while waiting for resource Issue in detail:- I am using the parallel ORACLE enterprise stage in my load job to load the transformed data. During load I am facing de...
by prasannak
Wed May 14, 2008 9:18 am
Forum: General
Topic: trigger condition
Replies: 11
Views: 9177

Thanks to Minhajudin and Craig... Now, I have more than one idea to play with... The reason why I detest cross application dependancies at scheduler level is the very very tight dependancy it creates on more than one application...but, some times, to keep things simple, this is the way to go...and i...
by prasannak
Tue May 13, 2008 10:40 am
Forum: General
Topic: trigger condition
Replies: 11
Views: 9177

Thanks Criag, I guess you are bent on not answering my question! :roll: :wink: I do not unfortunately participate in the decision making process of requirements... I just try my best to abide by the decisions...and if they are silly, then I question them...but, this requirement is not something ridi...
by prasannak
Tue May 13, 2008 10:39 am
Forum: General
Topic: trigger condition
Replies: 11
Views: 9177

Thanks Criag, I guess you are bent on not answering my question! :roll: :wink: I do not unfortunately participate in the decision making process of requirements... I just try my best to abide by the decisions...and if they are silly, then I question them...but, this requirement is not something ridi...
by prasannak
Tue May 13, 2008 9:06 am
Forum: General
Topic: trigger condition
Replies: 11
Views: 9177

thanks for all the responses... I have thought about doing what Craig had mentioned... But, its a requirement that enforces us not to have a cross application dependancy at the scheduler level directly... text file option has also been thought about...i know we can use the WaitforfileActivity...but,...
by prasannak
Mon May 12, 2008 9:27 pm
Forum: General
Topic: trigger condition
Replies: 11
Views: 9177

Nobody got any ideas :roll: :roll: :?
by prasannak
Mon May 12, 2008 7:26 am
Forum: General
Topic: trigger condition
Replies: 11
Views: 9177

trigger condition

I have a question regarding polling within datastage. The specific criteria being that we are required to trigger our datastage jobs(a master sequence rather) based on the successful completion of certain jobs from an existing application that is built using oracle as the database. Our master sequen...
by prasannak
Fri Apr 25, 2008 2:42 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: EtlStats question
Replies: 20
Views: 9120

Thanks a lot for the response... I have been able to progress a lot based on the inputs... However, I have more questions... I have read in your other posts that: You do not have to call DSJobReportDbDriver at all if you want. You can run DSJobReportDb directly. DSJobReportDbDriver was created so at...
by prasannak
Wed Apr 23, 2008 3:40 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: EtlStats question
Replies: 20
Views: 9120

Hi , I had a some questions on etlstats: I had imported the dsx file. 1) The install.txt specifies creating files in TCL prompt. Are these directories created under unix etl server? Our environment is usually constrained to allow creating additional folders since a project is allocated at a minimum ...
by prasannak
Wed Apr 23, 2008 8:58 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: EtlStats question
Replies: 20
Views: 9120

Suppose I do not want to install the whole package of etlstats to my project, would i be still able to leverage just the dsjobreportDB features of inserting stats into the table...? I am interested in the stats that are in etl_job_hist (start,end times, status etc)... I guess what i am trying to get...