DSXchange

What exactly do you want to accomplish by connecting to DS using K shell. If it is simply job maintenance (running, stoppnig, getting log, resetting, etc.) then that can be done using DSJOB. Is it all you want to do?

Regards,
Sumit

Yaa, and I guess Hash File always takes the last record.

-Sumit

If I understood it right in Output Link 2 you are making a lookup to a hash file which is being populated in Output Link 1. Now in this kind of a scenario the transformer that makes a lookup to this hash file will always wait for the hash file to get populated completely before processing the record...

Great chulet, I never knew about this option.

Thanks,
Sumit

What I would suggest is this: 1) While creating our hash file check the "Universe Stage Compatibility" option 2) In the main job make a look up to this hash file using a Universe Stage. This is because Universe Stage supports Relational operator in Lookup join conditions. Say the Universe ...

Thanks a lot Ketfos. It works. I was just wondering why don't we see all the columns when we do a "Select * From" and hash file in Command Interface.

Eg. - Try running SELECT * FROM DS_JOBOBJECTS in Command Interface. It won't return all the columns

Regards,
Sumit

Hi All,

Is there any command to describe a Hash File (like Describe of Oracle).

A command that expects the hash file name and when executed returns all the columns present in the hashfile with their datatypes.

Regards,
Sumit

I have had exactly the similar kind of a requirement before but couldn't find out a very good way of doing it. I guess the DataStage gurus will definitely have a better suggestion to make but here is what I did: 1) Get a list of all your user defined hash files in a sequential file - say A.txt. 2) e...

To find out the culprit records you can use constraints in the transformer on key columns. Push all the records with Null Key values to a sequential file and the one with Not Null Key values to the target hash file. This should help you in debugging.

Regards,
Sumit

This is because in your new project the file SDKSequences does not exist. What you did is one way of doing it but by diong this I have faced some issues in the past. A better way would be to use and of the SDK routines to create any Dummy sequence. The routine internally would create the file SDKSeq...

Roy can you tell me in which pdf I have to look for it. I couldn't find it in servjdev.pdf under the section called Stage Development Kit, Command Line Interface.

Regards,
Sumit

Thanky Roy. You are right. The server import is for the executables only.

Regards,
Sumit

Thanks a lot guys. This really helps.

Hi,

I want to use dsjob for triggering the jobs in a DS Server installed in UNIX box. Now dsjob.exe is a server side component. How can I use dsjob from an NT system(that has only client installed in it) to trigger a DataStage job in UNIX server. Is it possible at all?

Regards,
Sumit

Hi, I have three questions: 1) I want to automate the backup process using CLI commands. I see two commads available for triggering an export process - dsexport and dscmdexport. Can someone please tell me the difference between these two. 2) Can I trigger an export process using dsexport/dscmdexport...