Trace DBM1 Virtual Storage activity with RMF

Hello

Trace DBM1 Virtual Storage activity with RMF
Dear all ,

Reading an article (from Z/journal) about tracing the VS using RMF , i try
to test my reading , and get the following report :


(i extracted only the essential)

-
PRIVATE AREA SUMMARY

JOB NAME - GR0ZDBM1
PROGRAM NAME - DSNYASCP

REGION REQUESTED 5332K
REGION ASSIGNED (BELOW 16M) 5332K
REGION ASSIGNED (ABOVE 16M) 32.0M


LSQA/SWA/229/230
PAGES ALLOCATED
(IN BYTES) Above 16M 443M
USER REGION
PAGES ALLOCATED Above 16M 30.4M

The Region requested / assigned correspond exactly to the REGION= of
GR0ZDBM1


I've read some threads about Virtual Storage on the list, and understand
that Region size on the jcl does not limit the memory requirement of DB2,
as the region size controls only the use of Private Region Storage.

What the article said is Free VS = Region assigned (above 16M) -
(LSQA/SWA/229/230 allocacted + User Region allocated )
In my example, the result is negative ! how to interpret this ?

Should i change Region=0K in order to have the correct number ? IBM
recommends it but some threads warns about system down with 0K assigned ...

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Richard Humphris

Re: Trace DBM1 Virtual Storage activity with RMF
(in response to Hello)
Hi Dan,

I suspect your db2 sub-system is using dataspaces and/or hiperpools...
and so a lot of the 443M is not actually in the private address space of
GROZDBM1.

Rich Humphris
CNA

-----Original Message-----
From: DB2 Data Base Discussion List [mailto:[login to unmask email] On
Behalf Of Daniel Cremieux
Sent: Wednesday, February 18, 2004 8:10 AM
To: [login to unmask email]
Subject: Trace DBM1 Virtual Storage activity with RMF


Dear all ,

Reading an article (from Z/journal) about tracing the VS using RMF , i
try
to test my reading , and get the following report :


(i extracted only the essential)

-
PRIVATE AREA SUMMARY

JOB NAME - GR0ZDBM1
PROGRAM NAME - DSNYASCP

REGION REQUESTED 5332K
REGION ASSIGNED (BELOW 16M) 5332K
REGION ASSIGNED (ABOVE 16M) 32.0M


LSQA/SWA/229/230
PAGES ALLOCATED
(IN BYTES) Above 16M 443M
USER REGION
PAGES ALLOCATED Above 16M 30.4M

The Region requested / assigned correspond exactly to the REGION= of
GR0ZDBM1


I've read some threads about Virtual Storage on the list, and understand
that Region size on the jcl does not limit the memory requirement of
DB2,
as the region size controls only the use of Private Region Storage.

What the article said is Free VS = Region assigned (above 16M) -
(LSQA/SWA/229/230 allocacted + User Region allocated )
In my example, the result is negative ! how to interpret this ?

Should i change Region=0K in order to have the correct number ? IBM
recommends it but some threads warns about system down with 0K assigned
...

------------------------------------------------------------------------
---------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and
home page at http://www.idugdb2-l.org/archives/db2-l.html. From that
page select "Join or Leave the list". The IDUG DB2-L FAQ is at
http://www.idugdb2-l.org. The IDUG List Admins can be reached at
[login to unmask email] Find out the latest on IDUG conferences
at http://conferences.idug.org/index.cfm

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Martin Packer

Re: Trace DBM1 Virtual Storage activity with RMF
(in response to Richard Humphris)
A slight modification if I may, Richard. That 443 is what's left AFTER
you've taken dataspace out. Likewise hiperpools.

At 443MB I'd be disinclined to worry about DB2 Virtual Storage.

Having got everyone sensitised to the issue we are obviously now in the
phase of calming people down. :-)

Cheers, Martin

Martin Packer, MBCS Martin Packer/UK/IBM
020-8832-5167 in the UK (+44) (MOBX 273643, Internal 7-325167, Mobile
07802-245584)



"Humphris,Richard
P."
<Richard.Humphris To
@CNA.COM> [login to unmask email]
Sent by: DB2 Data cc
Base Discussion
List Subject
<[login to unmask email] Re: Trace DBM1 Virtual Storage
ORG> activity with RMF


18-02-04 20:17


Please respond to
DB2 Database
Discussion list
at IDUG






Hi Dan,

I suspect your db2 sub-system is using dataspaces and/or hiperpools...
and so a lot of the 443M is not actually in the private address space of
GROZDBM1.

Rich Humphris
CNA

-----Original Message-----
From: DB2 Data Base Discussion List [mailto:[login to unmask email] On
Behalf Of Daniel Cremieux
Sent: Wednesday, February 18, 2004 8:10 AM
To: [login to unmask email]
Subject: Trace DBM1 Virtual Storage activity with RMF


Dear all ,

Reading an article (from Z/journal) about tracing the VS using RMF , i
try
to test my reading , and get the following report :


(i extracted only the essential)

-
PRIVATE AREA SUMMARY

JOB NAME - GR0ZDBM1
PROGRAM NAME - DSNYASCP

REGION REQUESTED 5332K
REGION ASSIGNED (BELOW 16M) 5332K
REGION ASSIGNED (ABOVE 16M) 32.0M


LSQA/SWA/229/230
PAGES ALLOCATED
(IN BYTES) Above 16M 443M
USER REGION
PAGES ALLOCATED Above 16M 30.4M

The Region requested / assigned correspond exactly to the REGION= of
GR0ZDBM1


I've read some threads about Virtual Storage on the list, and understand
that Region size on the jcl does not limit the memory requirement of
DB2,
as the region size controls only the use of Private Region Storage.

What the article said is Free VS = Region assigned (above 16M) -
(LSQA/SWA/229/230 allocacted + User Region allocated )
In my example, the result is negative ! how to interpret this ?

Should i change Region=0K in order to have the correct number ? IBM
recommends it but some threads warns about system down with 0K assigned
...

------------------------------------------------------------------------
---------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and
home page at http://www.idugdb2-l.org/archives/db2-l.html. From that
page select "Join or Leave the list". The IDUG DB2-L FAQ is at
http://www.idugdb2-l.org. The IDUG List Admins can be reached at
[login to unmask email] Find out the latest on IDUG conferences
at http://conferences.idug.org/index.cfm

---------------------------------------------------------------------------------

Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home
page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select
"Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org
. The IDUG List Admins can be reached at [login to unmask email]
Find out the latest on IDUG conferences at
http://conferences.idug.org/index.cfm

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

nguyen duc tuan

Re: Trace DBM1 Virtual Storage activity with RMF
(in response to Martin Packer)
Hi Daniel

Region size does not control all areas used by DB2 , as region parameter
controls only the 'low private storage ' , Most of DB2 memory usage are
in 'High private '
For better understanding about these differents storage area , have a look
at the Z/OS MVS Programming Services Guide , chapTer Virtual Storage
management

Regards

Duc

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Bob Jeandron

DBM1 storage
(in response to nguyen duc tuan)
Have been monitoring storage via RMF, VSTOR(D,xxxxDBM1), for DBM1 address
space and found that we are using about 15 to 20M each day. It seems that
there is a compression at some point but that seems to just be buying a
little time. It is difficult to see how DB2 can stay up continuously
without recycle to clear up the creeping storage. We started monitoring
this because of a short on storage condition in October which brought down
the subsystem. We have applied PTFs to resolve that particular problem,
but can see other problems down the line with this storage creep.

Address space limit approx. 1600M
Current MAX LSQA/SWA/229/230 PAGES ALLOCATED and MAX USER
REGION PAGES ALLOCATED
approx. 1000M
Storage cushion approx. 100M
Without considering storage compression
approx. 25 days before hit storage cushion...
(600M-100M)/20M

Haved reviewed John Campbell's excellent presentation,
http://www-306.ibm.com/software/data/db2/os390/techdocs/DMDB13.pdf,
and it answered a lot of ?s, but is this 15 to 20M a problem?
Anyone else see similar storage usage in DBM1?
How often do you recycle DB2?

Thanks in advance.

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Martin Packer

Re: DBM1 storage
(in response to Bob Jeandron)
Turn on Stats Trace Class 6 (IFCID 225) and tell us which area is the
growing one.

Thanks, Martin

Martin Packer, MBCS CITP Martin Packer/UK/IBM
020-8832-5167 in the UK (+44) (MOBX 273643, Internal 7-325167, Mobile
07802-245584)

"Borrowing your watch and using it to tell you which way South is"

Blog:
http://[login to unmask email]



Bob Jeandron
<[login to unmask email]
O.COM> To
Sent by: DB2 Data [login to unmask email]
Base Discussion cc
List
<[login to unmask email] Subject
ORG> DBM1 storage


18-12-04 13:57


Please respond to
DB2 Database
Discussion list
at IDUG






Have been monitoring storage via RMF, VSTOR(D,xxxxDBM1), for DBM1 address
space and found that we are using about 15 to 20M each day. It seems that
there is a compression at some point but that seems to just be buying a
little time. It is difficult to see how DB2 can stay up continuously
without recycle to clear up the creeping storage. We started monitoring
this because of a short on storage condition in October which brought down
the subsystem. We have applied PTFs to resolve that particular problem,
but can see other problems down the line with this storage creep.

Address space limit approx. 1600M
Current MAX LSQA/SWA/229/230 PAGES ALLOCATED and MAX USER
REGION PAGES ALLOCATED
approx. 1000M
Storage cushion approx. 100M
Without considering storage compression
approx. 25 days before hit storage cushion...
(600M-100M)/20M

Haved reviewed John Campbell's excellent presentation,
http://www-306.ibm.com/software/data/db2/os390/techdocs/DMDB13.pdf,
and it answered a lot of ?s, but is this 15 to 20M a problem?
Anyone else see similar storage usage in DBM1?
How often do you recycle DB2?

Thanks in advance.

---------------------------------------------------------------------------------

Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home
page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select
"Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org
. The IDUG List Admins can be reached at [login to unmask email]
Find out the latest on IDUG conferences at
http://conferences.idug.org/index.cfm

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Hans Selman

Re: DBM1 storage
(in response to Martin Packer)
On Sat, 18 Dec 2004 07:57:41 -0600, Bob Jeandron <[login to unmask email]>
wrote:

>Have been monitoring storage via RMF, VSTOR(D,xxxxDBM1), for DBM1 address
>space and found that we are using about 15 to 20M each day. It seems that
>there is a compression at some point but that seems to just be buying a
>little time. It is difficult to see how DB2 can stay up continuously
>without recycle to clear up the creeping storage. We started monitoring
>this because of a short on storage condition in October which brought down
>the subsystem. We have applied PTFs to resolve that particular problem,
>but can see other problems down the line with this storage creep.
>
>Address space limit approx. 1600M
>Current MAX LSQA/SWA/229/230 PAGES ALLOCATED and MAX USER
> REGION PAGES ALLOCATED
> approx. 1000M
>Storage cushion approx. 100M
>Without considering storage compression
> approx. 25 days before hit storage cushion...
> (600M-100M)/20M
>
>Haved reviewed John Campbell's excellent presentation,
>http://www-306.ibm.com/software/data/db2/os390/techdocs/DMDB13.pdf,
>and it answered a lot of ?s, but is this 15 to 20M a problem?
>Anyone else see similar storage usage in DBM1?
>How often do you recycle DB2?
>
>Thanks in advance.
>
>--------------------------------------------------------------------------
-------
>Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and
home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page
select "Join or Leave the list". The IDUG DB2-L FAQ is at
http://www.idugdb2-l.org. The IDUG List Admins can be reached at DB2-L-
[login to unmask email] Find out the latest on IDUG conferences at
http://conferences.idug.org/index.cfm

Bob

I saw your question, and we were also bitten by storage abends in the
begin of this year. In the presentation you mentioned are some
remarks "How much is enough?"
about spare space > 500 Mb (green).
There was our problem, so please checkout this very important figure.
After adjustments of our bufferpools we didn't have any storage problems.
From now on I'll allways check this spare space after ZPARM changes.

Hans Selman

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

Isaac Yassin

Re: DBM1 storage
(in response to Hans Selman)
Look for the RSU parameter (system area not DB2 ). If it's too high it can kill you, especially with expanded storage usage.


Isaac Yassin


-----Original Message-----
From: DB2 Data Base Discussion List [mailto:[login to unmask email] On Behalf Of Hans Selman
Sent: Thursday, December 23, 2004 1:06 AM
To: [login to unmask email]
Subject: Re: DBM1 storage

On Sat, 18 Dec 2004 07:57:41 -0600, Bob Jeandron <[login to unmask email]>
wrote:

>Have been monitoring storage via RMF, VSTOR(D,xxxxDBM1), for DBM1
>address space and found that we are using about 15 to 20M each day. It
>seems that there is a compression at some point but that seems to just
>be buying a little time. It is difficult to see how DB2 can stay up
>continuously without recycle to clear up the creeping storage. We
>started monitoring this because of a short on storage condition in
>October which brought down the subsystem. We have applied PTFs to
>resolve that particular problem, but can see other problems down the line with this storage creep.
>
>Address space limit approx. 1600M
>Current MAX LSQA/SWA/229/230 PAGES ALLOCATED and MAX USER
> REGION PAGES ALLOCATED
> approx. 1000M
>Storage cushion approx. 100M
>Without considering storage compression
> approx. 25 days before hit storage cushion...
> (600M-100M)/20M
>
>Haved reviewed John Campbell's excellent presentation,
>http://www-306.ibm.com/software/data/db2/os390/techdocs/DMDB13.pdf,
>and it answered a lot of ?s, but is this 15 to 20M a problem?
>Anyone else see similar storage usage in DBM1?
>How often do you recycle DB2?
>
>Thanks in advance.
>
>-----------------------------------------------------------------------
>---
-------
>Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and
home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at
http://www.idugdb2-l.org. The IDUG List Admins can be reached at DB2-L- [login to unmask email] Find out the latest on IDUG
conferences at http://conferences.idug.org/index.cfm

Bob

I saw your question, and we were also bitten by storage abends in the begin of this year. In the presentation you mentioned are some
remarks "How much is enough?"
about spare space > 500 Mb (green).
There was our problem, so please checkout this very important figure.
After adjustments of our bufferpools we didn't have any storage problems.
From now on I'll allways check this spare space after ZPARM changes.

Hans Selman

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html.
From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be
reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm

---------------------------------------------------------------------------------
Welcome to the IDUG DB2-L list. To unsubscribe, go to the archives and home page at http://www.idugdb2-l.org/archives/db2-l.html. From that page select "Join or Leave the list". The IDUG DB2-L FAQ is at http://www.idugdb2-l.org. The IDUG List Admins can be reached at [login to unmask email] Find out the latest on IDUG conferences at http://conferences.idug.org/index.cfm