Re: RPG and BLOCK(*YES)

[rob@xxxxxxxxx]

<snip>
If you have a reasonable assurance that the data is in key order -AND- the
blocking factor will retrieve the records that you DO want.  Although I
never
depend on this, cache seems to work fine here.
<endsnip>

Goes to show you why a RGZPFM KEYFILE(*FILE) can help performance.  If I
used a I,F and then dropped the key off of my F spec, and ran the RGZPFM
prior, then a difference in runtime of 44 minutes would have been achieved.
Granted, a RGZPFM of a 18m record file might take a few minutes.  But if
the report is an end of period report anyways, this might be a good time to
do some cleanup.

All tests were done on an 820-25BC running at 68% of dasd, % CPU used . . .
. . . . :       21.9.  This last figure was a snapshot.

Rob Berendt

==================
"They that can give up essential liberty to obtain a little temporary
safety deserve neither liberty nor safety."
Benjamin Franklin