Migrating hash file

dr.murthy · Post by **dr.murthy** » Wed Apr 28, 2010 2:14 am

Hi,
In my project some of the jobs we are going to migrate server edition to px,we have an issue with hash file while migrationg we are aditinally building a one serverjob and one parallel job new , which loads hash file data into seq file then seq file into dataset.

is there any other way to implement this changes to get hash file data on parallel console.

ray.wurlod · Post by **ray.wurlod** » Wed Apr 28, 2010 4:28 am

No. The Hashed File stage is only available in server jobs or server shared containers. You *can* use server shared containers in parallel jobs under specific conditions, which are detailed in Chapter 2 of the Parallel Job Developer's Guide.

PhilHibbs · Post by **PhilHibbs** » Thu Jul 29, 2010 6:25 am

Is there a Parallel Job stage that gives equivalent functionality to the Hashed File? Lookup File Set doesn't work because writing to it clears down existing content (unless I'm missing something)

Thinking a little further - I'm half-expecting the answer to be "no", given the parallel and partitioned nature of the task. Maybe I can manage by re-creating the LFS with the additional content at the end of each run.

*Update*: Hm. Fly in the ointment - can't stream out from a Lookup File Set in order to create a new set with the additional rows. Any suggestions, other than also creating a regular File Set at the same time as the lookup File Set?

ArndW · Post by **ArndW** » Thu Jul 29, 2010 9:07 am

Hashed files can be accessed as ODBC sources - performance is going to be quite a bit slower than accessing them in server jobs.