RE: How to optimize COPY *ADD

[thomas@xxxxxxxxxxx]

Dimitri:

Please post back here with the results of your parallel-CPYF trial. I'm sure 
others besides me will be interested. If the index updates are the big problem, 
I'm not sure that parallel CPYFs will have much effect, but it's worth knowing.

I'd personally go with the suggestions to use either MAINT(*REBLD) or RMVM and 
set them back after the CPYF finishes. And then I'd definitely run the ADDMs in 
parallel. I have a RGZDBRPFM (Reorganize by Database Relations Physical File 
Member) tool that does the same sequence except with a RGZPFM where you'd have 
a CPYF. When running against some massive BPCS databases, the reorgs were cut 
from well over 20 hours down to under two hours. The whole key there was in 
managing the indexes.

Tom Liotta

On Fri, 27 July 2001, Dimitri Efimov wrote:

> 
> That's correct, we are copying daily built file to history file, the point
> is - the history file has more than 200 mln. of records, even more - it has
> 18 index files. So it's impossible to find "keyed sequence". 
> 
> But I'm going to try the solution Buck's solution ("Multiple CPYF commands
> in parallel might be better if you have a good way to partition the records
> such as FROMRCD(1)
> TORCD(1000000)...FROMRCD(1000001) TORCD(2000000) ...) hope it will work. 

-- 
Tom Liotta
The PowerTech Group, Inc.
19426 68th Avenue South
Kent, WA 98032
Phone  253-872-7788
Fax  253-872-7904
http://www.400Security.com


___________________________________________________
The ALL NEW CS2000 from CompuServe
 Better!  Faster! More Powerful!
 250 FREE hours! Sign-on Now!
 http://www.compuserve.com/trycsrv/cs2000/webmail/




+---
| 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
+---