CPYF Performance

[HBMB]

Hi,

Anywhere to speed up CPYF performance if the to file has multiple logical
files that are defined with unique access path? Anyone has any other
alternative to CPYF method to copy a large files with 1.8Mil records?

Thanks.

Mee Ling

[Thomas]

Mee Ling:

For most problems of this type (RGZPFM is another), the answer is to
avoid the index rebuilds.

For some logicals this can be done by a CHGLF to change MAINT(*REBLD)
before the CPFY and resetting MAINT() to its original value after the
CPYF -- note that RECOVER() might also need setting/resetting. For
others, you can do a RMVM before the CPYF and an ADDLFM after the CPYF.
In some circumstance, you can CRTDUPOBJ for the physical (without data)
and the logicals into another library, then delete the LFs before the
CPYF and CRTDUPOBJ them back after the CPYF.

I suspect there are other ways, but however you do it, just remember
that the definitions should be the same after as they were before and
that the CPYF by itself generally isn't where the problem is.

Tom Liotta

In article <7t1b7h$doh$1...@news5.jaring.my>,

--
Tom Liotta
AS/400 systems programmer

Sent via Deja.com http://www.deja.com/
Before you buy.

[Krokkie]

Also do the cpyf from record number 1 , speeds it up to

PW Botha

Thomas wrote in message <7t1g2e$92m$1...@nnrp1.deja.com>...

[Bob Frantz]

You can also remove the logical file members before the CPYF ((RMVM), then
add the logical file members back after the CPYF (ADDLFM). This is not the
same as deleting the logical files. You don't need the DDS source to add a
logical file member.

--
Bob Frantz
bo...@rkfrantz.com

RK Frantz Software Inc
http://www.rkfrantz.com

HBMB wrote in message <7t1b7h$doh$1...@news5.jaring.my>...

[Thomas]

And always remember that the member you ADD must have the same
attributes as the one you RMV'd. A very particular item to be careful of
is the DTAMBRS() attribute. If an LF is built over multiple data
members. you need to track those for later reADDing.

Tom Liotta

In article <7t26i0$jfa$1...@nntp2.atl.mindspring.net>,

--

[Koos Bezemer]

Well,

I don't know what you can do about the to-file, but the from file of course
you can speed up with blocking (to 128K via specifying no of records to be
read as a calculation between 128K and record size) and reading sequential
(if the file is keyed, specify to start reading from record 1 to make it
read sequential).

Is the to-file itself sequential ?

For the to-file, specify at least that the force-to-write ratio is *NONE via
a chgpf command.

Koos

HBMB <hbm...@po.jaring.my> wrote in message
news:7t1b7h$doh$1...@news5.jaring.my...