Key Length in Hashed File

Post questions here relative to DataStage Server Edition for such areas as Server job design, DS Basic, Routines, Job Sequences, etc.

Moderators: chulett, rschirm, roy

Post Reply
Vinodanand
Premium Member
Premium Member
Posts: 112
Joined: Mon Jul 11, 2005 7:54 am

Key Length in Hashed File

Post by Vinodanand »

Hi,

This is in Correlation to my earlier post but thought it would be helpful to have it seperately.

i) My Hashed file would approximately have 2.5 million records and 67 columns and it was taking ever to write.So I split it into 6 Hashed Files. One thing I have noticed is that my key size (key columns in the Hashed file) are 24 bytes,when I reduce it to 18 bytes they seem to be a bit quicker. Is the key size proportional to the time it takes to build the hashed file.

I did not tune all the parameters and left the default values as it is .

Type :T30 File
Modulus :1
Group Size : 1
Split : 80
Merge : 20 (changed the parameter)
Record Size : 400(changed the parameter)
Large Record :1628

ii) One more thing i noticed is that the same job ran in 25 mins, i believe it is because of the load on the DS server. Is there a way I can check the same.

Thanks,
Vinod
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

The limiting factor on key size is the number of characters in the key, rather than the number of columns in the key. By default this limit is 768 characters. It can be tuned in uvconfig. There are substantial implications in increasing it, however, for example GROUP.SIZE 1 may no longer be appropriate.

The size of the key ought to make no discernable difference between 18-byte and 24-byte keys.

The only way to test your theory about machine load is to test under varying load conditions.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
ray.wurlod
Participant
Posts: 54607
Joined: Wed Oct 23, 2002 10:52 pm
Location: Sydney, Australia
Contact:

Post by ray.wurlod »

The limiting factor on key size is the number of characters in the key, rather than the number of columns in the key. By default this limit is 768 characters. It can be tuned in uvconfig. There are substantial implications in increasing it, however, for example GROUP.SIZE 1 may no longer be appropriate.

The size of the key ought to make no discernable difference between 18-byte and 24-byte keys.

The only way to test your theory about machine load is to test under varying load conditions.
IBM Software Services Group
Any contribution to this forum is my own opinion and does not necessarily reflect any position that IBM may hold.
Post Reply