More curiosity than anything. That seemed to be a drastic change in a number 
and I was wondering if it was normal or an indicator of a potential problem. I 
suspect it was the change in algorithm. Also, we have seen the number of TPF 
users increase by over 25% since we removed the 2G constraint. In any event, as 
long as page space occupancy doesn't go too much over 50% occupancy, I don't 
think I will worry about it.

We bit that same page packs bullet before we threw money at the problem and 
jumped from 16G to 56G to eliminate dasd paging. Our numbers under 4.4 were 
about 50% before the storage increase. When the storage was later increased to 
56G, the dasd usage was reduced to nearly 0 and practically all paging was 
between main and XSTORE. Now, we also see some dasd paging, but it is not 
anywhere close to a bothersome number.   


Regards,
Richard Schuh


> -----Original Message-----
> From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED]
> Behalf Of Alan Ackerman
> Sent: Wednesday, August 09, 2006 9:47 PM
> To: IBMVM@LISTSERV.UARK.EDU
> Subject: Re: External Page Usage
> 
> 
> We ran into something like this when we went from 4.4 to 5.1. 
> We actually=
>  ran out of page space. IBM 
> pointed us to a page in a manual that said we should have 
> enough page spa=
> ce for the sum of the size 
> all our virtual machines. We "Bit" the bullet and got more 
> page packs. I =
> don't remember details, but 
> there were algorithm changes in 5.1. Once we added the page 
> packs, I don'=
> t remember any actual 
> problems with it.
> 
> Do you really have a problem, or do you just not like the 
> numbers you are=
>  seeing?
> 
> On Wed, 9 Aug 2006 13:49:00 -0700, Schuh, Richard 
> <[EMAIL PROTECTED]> wrote=
> :
> 
> >With our current configuration - 36GB main, 26GB XSTORE, 19 
> 3390-03s for=
>  paging, we saw very 
> low numbers (< 10%) when we entered Q ALLOC PAGE while 
> running 4.4. Now, =
> on 5.2, the results of 
> the query show about  50% of the space is in use. Is this normal? 
> >
> >Regards,
> >Richard Schuh
> >=========================
> ==========================
> ============
> ===========
> 

Reply via email to