DSXchange

Could you specify exactly what you mean when

the entire database is geeting down because of this Job

? The only possible issue I know is when using the load method you might get the table, and perhaps the tablespace, into a "load pending" or "backup pending" state.

Sure, if you have the appropriate unzip program on UNIX, or else unzip it on the PC and FTP it over, or perhaps you might have software that lets you mount a NTFS network drive from your UNIX flavor. The unzip or other preprocessing can be done as a call from your sequential file stage once you've g...

So, I dont require to enable the multiple instances option, if I have to run the jobs sequentially and calling the job twice in between the job flow. Wrong. As stated above, at v5 you cannot run multinstance jobs without the Axcel pack. You can copy your Job1 into Job1a, and then call that as part ...

All DataStage job files are relative to the server and not the client. You can execute command-line unzip program a part of a sequential file stage, but it must be visible from the Server.

It isn't easy to corrupt a hashed file, and quite difficult to do so on a regular basis in normal production. In addition to posting the errors in the file, do you specify anything non-default in the file creation attributes? (i.e. is it a dynamic file?)

You need to use the DSAttachJob( ) function to return a file handle for each of the jobs you want to call. Since the file handle is part of the DSSetParam() function call you need to specify these parameter settings for each of the jobs you wish to call. If you recursively call the ControllerJob1 yo...

You have neglected to tell the aggregator stage that your incoming data is already sorted. Once you do so you will see the speed pick up and the stage will use almost no memory at all.

Kumar,

on AIX I have sorted many more records than that. I actually did call a UNIX sort from a DataStage job earlier today of about 48 million records.

There are a number of threads out there regarding the differences between Server and Parallel. The common designer canvas can misleas the new and unwary developer into thinking that the two products, Server and Parallel, differ mainly in their performance. This is a big mistake and an easy enough er...

M_keerthi2005,

you have now been given some possible solutions. What more do you need? You can pre-sort your data in UNIX and then feed to it DataStage for deduplication. You can sort and remove duplicate directly in the UNIX sort.

Sorting a mere 10 million records should be no problem. You might need to tell the sort command to use a different temporary directory if your /tmp is dimensioned rather small. Using sorted data on the aggregator stage will speed up the processing and also have the effect that the stage will use ver...

Hello Jim,

PX has a number of builtin functions for handling time and timestamps. If you have a timestamp value, you can use the function TimestampFromSecondsSince to subtract one second from a timestamp value.

This thread is an exact repeat of one that happened last week. If you have a reasonable limit on the output links (say 30) then this method will work efficiently. The easy alternate is what Anupam has suggested, by using an algorithm to loop around creating multiple "lines" by inserting <cr><lf> int...

I just noticed (when Ray mentioned the modify stage) that we are in the PX world and not in server anymore! Sorry about confusing the response.

And does the same error (or any error) happen when you call up the command manually from the TCL environment? Also, what if you change your SPOOL command to a "COMO ON {ComoFileName}" and "COMO OFF" - perhaps redirecting output to SPOOL causes some internal confusion in the job; turning on COMO shou...