Mary,

If you have sufficient paging space, you might consider using V-disk instead of 
T-disk. As Nick pointed out, they are not subject to fragmentation and the 
limits for the individual users and the system as a whole can be set and 
adjusted dynamically. If you are not already paging constrained and you have 
the space, it might be a way out of the fragmentation problem. Look under HELP 
CP DEFINE (RELATED and select VDISK. You may also need to set limits using SET 
VDISK.

We have been using V-disk in place of T-disk for the past several years and 
have never had a problem with it.

 -----Original Message-----
From:   VM/ESA and z/VM Discussions [mailto:[EMAIL PROTECTED]  On Behalf Of 
Mary Zervos
Sent:   Tuesday, January 24, 2006 6:43 AM
To:     [email protected]
Subject:        Re: Tdisk question

Thanks Jim for the insight.  I had forgotten about fragmentation.

Mary

Jim Bohnsack wrote:

> Mary--What you're seeing is fragmentation.  TDISK is allocated in one 
> whole piece to a user.  Think of PC disk allocation that you see when 
> you run a defrag.  If you have 1 allocated cylinder out in the middle 
> of an otherwise empty 100 cylinder TDISK space, you cannot allocate a 
> 99 cyl TDISK.  You would get that by the first user getting 50 
> cylinders and the 2nd user getting 1 cylinder.  When the first user 
> detaches the 50 cyl. disk, the 2nd user still has the 1 cyl in the 
> middle.  You are left with 99 cylnders in total free but you can only 
> allocate a max of 50 cylinders and 49 cylinders.
>
> Jim
>
> At 09:17 AM 1/24/2006, you wrote:
>
>> I have 500 cyl. of tdisk defined on one pack and 50 cyl. of tdisk 
>> defined on another.  One user used the 50 cyl. on the second pack  
>> and about 24 users have used 215 cyl.(43%used) on the first pack.  
>> Why is it that I can only allocate a 130 cyl. maximum tdisk on the 
>> first pack but can then go ahead and allocate a 110 cyl. tdisk, 
>> instead of being able to allocate *one* 240 cyl. tdisk?
>>
>> Thanks for any help.
>>
>> Mary Zervos
>> VM Systems Programmer
>> Binghamton University
>> [EMAIL PROTECTED]
>
>
> Jim Bohnsack
> Cornell Univ.
> (607) 255-1760

Reply via email to