accesing hashfile -doubt

[sravanthi]

Hi,

I have a job where i need to access a hashfile as lookup.

Will there be any affect when we access a hashfile created on dev project from testing project.Both dev and test projects are on same server.

I'm trying to load 3 lakh records to hash file.

Please let me know.

Thanks in advance!

[gateleys]

How does your team promote jobs from DEV to TEST and then to PROD environments?

[satya99]

It should not be problem, if you are accessing it.

Can you go a head and give more information.

[sravanthi]

I have generated hash file in dev account.

I promoted my code to test environment.I'm able to access the same hash which i have generated in Dev account.

Will there be performance impact when accesing a hashfile from dev account in Test .

[harishvardhan]

There will not be any performance issue if you load the hashed file in DEV and access from TEST. For Datastage only PATH & privileges to access the file matters ( It may be from DEV, TEST or PROD).

BUT
Better/Standard practice is

1. To make a "/test" directory in server.
2. Then run the job which creates hashed file from test environment
3. Use the hashed file form test environment.

[ray.wurlod]

How wide are the rows?
If 3 lakh * (rowsize) << 2GB then you are OK.
If not, you need to create your hashed file with 64-bit addressing.

[chulett]

What about when the job goes to Production? You still going to be accessing the hash in the Dev Account?

[sravanthi]

Ray,

i'm creating a file with 4 columns.

col1 VARCHAR2(15 ) ,
col2 CHAR(2 ),
col3 NUMBER(8),
col4 VARCHAR2(5 ).

Should i create 64 bit addressing ?

Chulett..in production we are using production files.

In test without making changes i have used the exported job which uses dev files.

[chulett]

Chulett..in production we are using production files.

Of course. I'm concerned about your hashed files, not your file files. My point is your assertion that the jobs are using account based (not pathed) hashed files in the dev project account. I can only see this continuing to work if your Production project is on the same server as everything else and even then it's not a very good idea. So you must be doing something else, or passing in a job parameter to use as the account name. If you've selected the option but not supplied an account name, then it is using the current account for hashed file storage. So while you may be using 'dev files' to populate them, you are using test hashed files.

Just trying to understand.

[ray.wurlod]

3 lakh * 50 bytes = 1.5MB only. You definitely don't need 64-bit addressing.