Hi All,
I have used Hashed file plugin in DataStage jobs. But I dont have any idea about CREATE.FILE <hash_file_name>. If i create a hashed file using the above syntax in Job Control, then
is it possible to access the same file using plugin (HashedFile Plugin)?
What is modulo?
Since I have used only plugins, I dont have idea about modulo, type of file (dynamic/static type2/3/.../18), number of hash buckets etc...
Also, If i use HashedFile Plugin in Jobs, where we can specify those values (modulo, number of hash buckets)?
Thanks,
Bala
Hashed File
Moderators: chulett, rschirm, roy
-
vmcburney
- Participant
- Posts: 3593
- Joined: Thu Jan 23, 2003 5:25 pm
- Location: Australia, Melbourne
- Contact:
If you are using DataStage 5 you can check the "Create File" check box in the plug in properties which ensures the hash file gets created whenever the job is run. You do not need to change any of the settings for the Hash file, it is created as a Dynamic hash file which in most cases will be okay.
If you are using older versions of DataStage you will find that the hash file is created for you when you validate your job. I've made the mistake in the past of adding a new hash file to a job and then trying to run it without a validation and getting an error. You need to validate once to create that hash file.
You should not need to run the CREATE.FILE command in job control code.
The only time you need to reconfigure your hash file and change the type or settings is when you have a hash file receiving a very large number of rows and you need to improve the performance of the jobs that create and use that hash file. A hash file is a Universe table and details about configuring them can be found in the administration manuals at http://www.informix.com/answers/english/puniv96.htm. I believe it is taught in the Advanced Datastage course.
regards
Vincent
If you are using older versions of DataStage you will find that the hash file is created for you when you validate your job. I've made the mistake in the past of adding a new hash file to a job and then trying to run it without a validation and getting an error. You need to validate once to create that hash file.
You should not need to run the CREATE.FILE command in job control code.
The only time you need to reconfigure your hash file and change the type or settings is when you have a hash file receiving a very large number of rows and you need to improve the performance of the jobs that create and use that hash file. A hash file is a Universe table and details about configuring them can be found in the administration manuals at http://www.informix.com/answers/english/puniv96.htm. I believe it is taught in the Advanced Datastage course.
regards
Vincent
For some detailed info on hashed files go to http://www.fitzlong.com and click on support and technical papers. You'll find some white papers from the early 90s (but still valid) pertaining to hashed files. It is geared towards the UniVerse user and is probably more info than a DS developer needs to know, but is nice to get the full understanding of hashed files.
In terms of mentioning type....while all the types are available to a DS user, typically the ETL developer only needs to use the dynamic (type 30).
In terms of mentioning type....while all the types are available to a DS user, typically the ETL developer only needs to use the dynamic (type 30).