Another thought .... jrf on jBASE 5.0 will use hash method 5 to 
calculate what groups the items go in to. This differs from 
jBASE 4.0 where it will use hash method 3. So you'll get a 
different distribution of items in the file. 

Maybe this differing hash method is causing the problems?

Could you try it by forcing the resize to be hash method # 3 
like this ?

   jrf -M3 FBNK.RE.CONSOL.PROFIT

This would need to be actually resized -- the -M3 and -Rv 
option together seems to confuse jrf

Greg
> 
> 
> Is it possible that that bucket size has been reduced for some reason  
> on jBase 5??
> 
> I've not been paying complete attention to this thread but have you  
> done a jstat before and after running jrf to see what it says??
> 
> Simon
> 
> ---------------------------------
> Simon Verona
> Director
> Dealer Management Services Ltd
> 
> Sent from my iPhone
> 
> On 11 May 2009, at 19:42, Jim Idle <[email protected]> wrote:
> 
> >
> > Christophe wrote:
> >> Hi Jim,
> >>
> >> Sorry I keyed the JBase 4 release incorrectly. Should be 4.1 (4.19)
> >> Patch 0758
> >>
> >> We have had not problem with the jBase 4.1 (T24 R05) so far... We 've
> >> logged a call with Temenos to explain the suddent increase in Modulo
> >> with jBase 5.0. But they are advising us that JBase 5.0 is correct...
> >> Which I cannot believe... Still got the call opened. Thanks for your
> >> feedback.
> >>
> > Then I would ask them the following:
> >
> > 1) If jBASE 5 is correct here, then should not the entire prior base  
> > be
> > patched to fix the previous jrf?
> > 2) Why is jBASE 5 correct in oversizing the file 3 or 4 times bigger
> > than it should be?
> >
> > I can only think that someone thought it a good idea to increase the
> > expansion margin by a factor of 3, which makes no sense. If this is
> > indeed what jrf on jBASE 5 should be saying, then the reason why needs
> > to be documented. Files that are too big are not generally a huge
> > overhead, but if you are doing SELECT and READNEXT a lot, then you  
> > will
> > have a lot of empty buckets to traverse and a lot of wasted file cache
> > doing it.
> >
> > Jim
> >
> > >
> 
> > 
> 




--~--~---------~--~----~------------~-------~--~----~
Please read the posting guidelines at: 
http://groups.google.com/group/jBASE/web/Posting%20Guidelines

IMPORTANT: Type T24: at the start of the subject line for questions specific to 
Globus/T24

To post, send email to [email protected]
To unsubscribe, send email to [email protected]
For more options, visit this group at http://groups.google.com/group/jBASE?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to