Re: File Reads

[VENU YAMAJALA <venu_yamajala@xxxxxxxxxxxx>]

Your case is somewhat different from what we are having. We have two datafiles.
We do a sequential read from one file, do some processing (like aging..etc) and
then write the record to the second file. Typically, the first file is input
only and the second one is output only files. These two thus are perfect
candidates for SEQONLY parameter.

Our record length is 118. So if we dont specify blocking factor, the system by
default takes 4096 buffer i.e., a blocking factor of approx 31 in our case. We
wanted to increase this so that we can process the job much faster and utilise
the CPU time more efficiently. We came up with the formula to calculate the
blocking factor for a file based on machine type and record length as

for RISC :    (64*1024) / RecLen
for CISC :   (32*1024) / RecLen

This way, we are getting a blocking factor of around 272. But I wanted to know
is there any other efficient formula that makes the job run faster? If we
increase the blocking factor, we get some potential paging problems. Is there
any work around to avoid these paging problems but still having a large
blocking factor?

Thanx in advance for any suggestions.

Rgds
Venu
+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---