View previous topic :: View next topic
|
Author |
Message |
space_admin
New User
Joined: 04 Apr 2006 Posts: 7
|
|
|
|
Hello together,
is there any reason NOT to release freespace from VSAM components? Here an example:
SPLITS-CI-------------93
SPLITS-CA--------------0
FREESPACE-%CI---------40
FREESPACE-%CA---------10
FREESPC-------2101248000
HI-A-RBA------2949120000
HI-U-RBA-------942243840
SPACE-TYPE------CYLINDER
SPACE-PRI-----------4000
SPACE-SEC------------300
What is a possible reason for this large freespace? When I calculate:
number ci-splits multiply with freespace per ci, I come to 2MB not to 2GB. That means in my opinion that the primary space is too large.
Now it is 1/3 full so why don't define a new cluster with 1/3 space and repro the data?
Are there any tools which release the space in a easier manner?
Kind regards,
space_admin |
|
Back to top |
|
|
martin9
Active User
Joined: 01 Mar 2006 Posts: 290 Location: Basel, Switzerland
|
|
|
|
hy space_admin,
i don't know any method to do this.
the FREESPACE %CI says only that this percentage
of space must be free or a control interval split will
automatically occur. the same for FREESPACE %CA,
what means analog to a CI, that this control area
automatically splits to fit the FREESPACE requirement.
if you think this is too large then minimize this value.
therefore create any new cluster with same size,
but with different FREESPACE values, and repro this.
but this decision is up to your choice...
martin9 |
|
Back to top |
|
|
space_admin
New User
Joined: 04 Apr 2006 Posts: 7
|
|
|
|
hi martin9,
Quote: |
hy space_admin,
i don't know any method to do this.
|
to do what?
I have a question to you:
For this dataset: what size is the minimum freespace of this data component?
Kind regards,
space_admin |
|
Back to top |
|
|
|