DSXchange

Hi, Ray once posted that for Code -99 there is no specific diagnostic handler within DataStage code. Try making a copy of that seq job which is giving you the error and if you still face the same error then make a copy of server job and try running it... And if it works fine then rename it...... Als...

Some times Hashed file size may be a problem.
Try using a 64 bit hashed file. This might solve your problem.

Because in a 64 bit hashed file largest address that can be represented is 19million TB.

Yes Of course. You can develop a scd 1 and scd 2 in server edition. Using OCI stage you have to deveop a source view and a deleted view(Which is target minus source in order to get deleted records). You have to add indicator flag which says whether it is current record or updated or deleted. And obv...

Simply you can add a transformer before hashed file and you can pass null values to the columns and as Ray mentioned you have check the option ''Delete file before create".

mshilpa wrote: I am trying to insert "Blank" in one of my column where my column data type is DATE.

If you are using any database like oracle then you have to give 'Y' in nullable column in table or view as well as in DS. Otherwise you might get "cannot insert null error".

Hi, In this case you can initially load the data with a bulk loader and then you can go for normal job design for incremental load(Update and delete). Also you can use link partitioner to load target table so that it can also reduce the time. Along with you can also use a Primary key indexes, useful...

how many dsx are you going to import?

Export all dsx once again and try.
(OR)
The job might be locked check Job resources in Director and if there is any lock release it if it is a existing job.
(OR)
And if you are going to import more number of jobs then the connection issue might also be a prob.

kandyshandy wrote: In your method, you don't need 5 seq files at all. You can directly load to hash files.

Yes you are right. I removed that and tried and it's working and faster too....

Thanks,
Kandy.

@rohit_mca2003: Hope this might be suffcient enough to get your desired output

Kandyshandy provided the right method. But in order to get your o/p in a simple way and with same order you have mentioned then you can have source and link partitioner with 5 seq after and 5 hash file and followed by a transformer where you can pull the data from the hash files and finally a seq fi...

"Input Data: ------------- Id Fname Sname Section Ques Ans -------------------------------------------------------------------- 100 Sam Roy S1 Q11 A11 100 Sam Roy S1 Q12 A12 100 Sam Roy S1 Q13 A13 100 Sam Roy S2 Q21 A21 100 Sam Roy S2 Q22 A22 I like output data as follows in one row and differe...

If you are using oracle then use a primary key index and if possible add a parallel hint to your query. So that you no need to change your DS code.

Hi,

Just try copying the installed files from a system where this OCI stage function is working properly and paste it into yours ie., files in C:\Program files\ Ascential\ Datastage 7.5.1 but don't re - install.

Even i faced this issue and got rectified this way.

Probably the Database settings won't be properly set in case if you are running those jobs for the first time.
Please check with the server.

Hi,

Probably job might be corrupted.

If you are running a job with high volume of data and when you stop the job may be the oracle might try to commit those rows. So the query might still run. In that case you can't avoid that.