Datasets

sshettar · Post by **sshettar** » Wed Aug 12, 2009 2:47 pm

Hi All,

I know datasets are very useful in terms of performance, but i was just curious to know if the datasets causes a problem if we decide to use different config file as compared to when cretaed it
As in lets say we are using 4 nodes and we create a dataset , just wondering if we can read the same dataset from a different job which is using a config file ( may be less or may be more number of nodes).

thanks in advance

chulett · Post by **chulett** » Wed Aug 12, 2009 3:10 pm

sshettar · Post by **sshettar** » Thu Aug 13, 2009 6:50 am

Thanks Craig!!!
but One small confusion , Is "YES" for the first part of my question or for my second part of my question.

i know this confusion is because of the way i framed the question. i have put two contradicting question out there. the first part says :does the dataset causes problem when the configuration file is changed.and my second part of the same question states" is it possible to read the dataset created with different config file in a job with different config file?
Hence would want to know for which part of my question is the answer "YES"

emma · Post by **emma** » Mon Aug 17, 2009 1:21 pm

data sets are platform and configuration specific

chulett · Post by **chulett** » Mon Aug 17, 2009 1:26 pm

"wondering if we can read the same dataset from a different job which is using a config file ( may be less or may be more number of nodes)."

It is perfectly valid to read a dataset created on X nodes by a job that is running on a different number of nodes. That's what I was "yes'ing".