EDM POOL CONCERN

Ken Hynes

EDM POOL CONCERN
This one is a "puzzler". We have an EDMPOOL that we have been working with.
It's Free Pages has on occasion gone as low as 2 or 3% -- no load failures
fortunately and overall hit ration performance looks good. I've always
thought that the guideline was to keep 10-15 % free wherever possible. We
recently adjusted the pool adding about 15% more pages and as of Tuesday
(the system cycles Sunday) were as low as 7%. I suspect Dynamic Cacheing is
the cuplrit, but don't know of a way to "Prove" it. Nor am I aware of any
decent means of "calculating" a good EDMPOOL size (other than the INSTALL
CLIST).

Here's the basic Layout of the Pool:

Full Fail 0 Free......... 1890
DBD.......... 1090
SKCT......... 57
CT........... 30
SKPT......... 20424
PT........... 4509
EDM Data Space Size 30000
Full Fail 0 Free......... 0
Dynamic prep. 30000
Efficiency Requests Loads
DBD........... 29217144 353
CT............ 1095149 48
PT............ 40784287 14151
Dynamic prep.. 9875654 1765457

NOTE MAXKEEPD is set at 5000

Anyone have any thoughts? Obviously we can continue to adjust the Pool and
add more storage, but it would be nice to have some idea as to how to get
to an "Acceptable" value -- again I'm working with trying to get 10-15%
Free at max usage since that's what I've understood as the "recommended"
level to try to maintain. One thing I have notioced in comparison with some
other systems I am more familiar with was the SKPT is much larger than I am
accustomed to seeing. Again any input based on previous experience with
this would be apreciated.

Regards.....Ken Hynes

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

William Favero

Re: EDM POOL CONCERN
(in response to Ken Hynes)
Assuming you are NOT running Version 8, I would think you could simply
assume a well utilized EDM Pool. If you are not planning any major
changes that could have a significant affect on the EDM Pool.

I was also strongly suggest moving your Dynamic Statement Cache to a
dataspace and completely take it out of the equation.

Willie

My DB2 blog --> http://blogs.ittoolbox.com/database/db2zos



Ken Hynes wrote:

>This one is a "puzzler". We have an EDMPOOL that we have been working with.
>It's Free Pages has on occasion gone as low as 2 or 3% -- no load failures
>fortunately and overall hit ration performance looks good. I've always
>thought that the guideline was to keep 10-15 % free wherever possible. We
>recently adjusted the pool adding about 15% more pages and as of Tuesday
>(the system cycles Sunday) were as low as 7%. I suspect Dynamic Cacheing is
>the cuplrit, but don't know of a way to "Prove" it. Nor am I aware of any
>decent means of "calculating" a good EDMPOOL size (other than the INSTALL
>CLIST).
>
>Here's the basic Layout of the Pool:
>
> Full Fail 0 Free......... 1890
> DBD.......... 1090
> SKCT......... 57
> CT........... 30
> SKPT......... 20424
> PT........... 4509
> EDM Data Space Size 30000
> Full Fail 0 Free......... 0
> Dynamic prep. 30000
> Efficiency Requests Loads
> DBD........... 29217144 353
> CT............ 1095149 48
> PT............ 40784287 14151
> Dynamic prep.. 9875654 1765457
>
>NOTE MAXKEEPD is set at 5000
>
>Anyone have any thoughts? Obviously we can continue to adjust the Pool and
>add more storage, but it would be nice to have some idea as to how to get
>to an "Acceptable" value -- again I'm working with trying to get 10-15%
>Free at max usage since that's what I've understood as the "recommended"
>level to try to maintain. One thing I have notioced in comparison with some
>other systems I am more familiar with was the SKPT is much larger than I am
>accustomed to seeing. Again any input based on previous experience with
>this would be apreciated.
>
> Regards.....Ken Hynes
>
>---------------------------------------------------------------------------------
>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

Chris White

Re: EDM POOL CONCERN
(in response to William Favero)
I agree with Willie and would like to expand a bit further...

I've never held with the conventional wisdom about maintaining "sufficient"
(pick any number) free space in the EDM pool. Just look at your hit
performance and fragmentation (frag is becoming less an issue). Your PT
ratio of physical to logical loads is very good... 99.9% of your access is
logical. It looks like your DBDs are loading ONCE and then staying in
cache (a GOOD THING). These numbers look excellent. Failures are ZERO.
Any effort to maintain additional free space in this pool (assuming these
stats are typical and the work load is stable) would just waste a valuable
resource.

Best regards,
Chris White
Recovering (ex)DB2 Sysprog
Caterpillar Inc.


These are just my opinions, not my employer's and blah, blah, blah...


===<snip>===>
On Tue, 13 Dec 2005 12:25:02 -0600, Willie Favero <[login to unmask email]>
wrote:

>Assuming you are NOT running Version 8, I would think you could simply
>assume a well utilized EDM Pool. If you are not planning any major
>changes that could have a significant affect on the EDM Pool.
>
>I was also strongly suggest moving your Dynamic Statement Cache to a
>dataspace and completely take it out of the equation.
>
>Willie
>
>My DB2 blog --> http://blogs.ittoolbox.com/database/db2zos
>
>
>
>Ken Hynes wrote:
>
>>This one is a "puzzler". We have an EDMPOOL that we have been working
with.
>>It's Free Pages has on occasion gone as low as 2 or 3% -- no load failures
>>fortunately and overall hit ration performance looks good. I've always
>>thought that the guideline was to keep 10-15 % free wherever possible. We
>>recently adjusted the pool adding about 15% more pages and as of Tuesday
>>(the system cycles Sunday) were as low as 7%. I suspect Dynamic Cacheing
is
>>the cuplrit, but don't know of a way to "Prove" it. Nor am I aware of any
>>decent means of "calculating" a good EDMPOOL size (other than the INSTALL
>>CLIST).
>>
>>Here's the basic Layout of the Pool:
>>
>> Full Fail 0 Free......... 1890
>> DBD.......... 1090
>> SKCT......... 57
>> CT........... 30
>> SKPT......... 20424
>> PT........... 4509
>> EDM Data Space Size 30000
>> Full Fail 0 Free......... 0
>> Dynamic prep. 30000
>> Efficiency Requests Loads
>> DBD........... 29217144 353
>> CT............ 1095149 48
>> PT............ 40784287 14151
>> Dynamic prep.. 9875654 1765457
>>
>>NOTE MAXKEEPD is set at 5000
>>
>>Anyone have any thoughts? Obviously we can continue to adjust the Pool and
>>add more storage, but it would be nice to have some idea as to how to get
>>to an "Acceptable" value -- again I'm working with trying to get 10-15%
>>Free at max usage since that's what I've understood as the "recommended"
>>level to try to maintain. One thing I have notioced in comparison with
some
>>other systems I am more familiar with was the SKPT is much larger than I
am
>>accustomed to seeing. Again any input based on previous experience with
>>this would be apreciated.
>>
>> Regards.....Ken Hynes
>>
>>--------------------------------------------------------------------------
-------
>>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
>>
>>
>>
>
>---------------------------------------------------------------------------
------
>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

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

Ken Hynes

Re: EDM POOL CONCERN
(in response to Chris White)
Hi,

Appreciate the feedback on the EDMPOOL concerns. Per Wille, I do
have 30000 pages allocated to Dynamic Cache and I know it's being utilized.
What prompted the question and concern was that this system apparently uses
a significant amount of KEEPDYNAMIC(YES) which according to the
Applications Programming guide caches the Prepare "Locally" -- the SQL
String would go to the Data Space Dynamic Cache. Sorry, I should have
included that in the original "Post". Per the DOC: "When the dynamic
statement cache is active, and you run an application bound with KEEPDYNAMIC
(YES), DB2 retains a copy of both the prepared statement and the statement
string." I am happy overall with the Hit Ratios as mentioned by Chris White
and perhaps I am being a bit "too careful" in trying to get a larger "Free
Shelf" in the Pool. There has been discussion of adding some new
Applications and files to the system which was why I was looking to
increase the Free Space. I am also wondering if the MAXKEEPD parm (set at
5000) is too high as STATS reports show "0" statements Discarded. We
increased the Pool by about 10% recently and went from 3% Free at MAx
Allocation to 7% Free -- so "Something" likes the new storage. Again, we
don't have a "Problem" and I appreciate the feedback. It "Aint Broke" as
Chris and WIllie point out so maybe move on to a more "immediate" problem.

Thanks.....Ken Hynes

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

William Favero

Re: EDM POOL CONCERN
(in response to Ken Hynes)
If your dynamic state cache is defined in a dataspace, it will not have an
affect on your EDM pool.

However, KEEPDYNAMIC YES in conjunction with MAXKEEPD will affect the DBM1
address space storage utilization. I would suggest taking a look at John
Campbell's presentation on storage management or giving his webcast on
storage management a listen.

Presentation URL
http://www.ibm.com/support/docview.wss?rs=64&context=SSEPEK&q1=campbell&uid=
swg27006985&loc=en_US&cs=utf-8&lang=en
Tiny URL - http://tinyurl.com/bsty8

Webcast replay information is in my blog entry -
http://blogs.ittoolbox.com/database/db2zos/archives/006846.asp

Willie

My DB2 blog --> http://blogs.ittoolbox.com/database/db2zos



Ken Hynes wrote:

>Hi,
>
> Appreciate the feedback on the EDMPOOL concerns. Per Wille, I do
>have 30000 pages allocated to Dynamic Cache and I know it's being utilized.
>What prompted the question and concern was that this system apparently uses
>a significant amount of KEEPDYNAMIC(YES) which according to the
>Applications Programming guide caches the Prepare "Locally" -- the SQL
>String would go to the Data Space Dynamic Cache. Sorry, I should have
>included that in the original "Post". Per the DOC: "When the dynamic
>statement cache is active, and you run an application bound with
KEEPDYNAMIC
>(YES), DB2 retains a copy of both the prepared statement and the statement
>string." I am happy overall with the Hit Ratios as mentioned by Chris White
>and perhaps I am being a bit "too careful" in trying to get a larger "Free
>Shelf" in the Pool. There has been discussion of adding some new
>Applications and files to the system which was why I was looking to
>increase the Free Space. I am also wondering if the MAXKEEPD parm (set at
>5000) is too high as STATS reports show "0" statements Discarded. We
>increased the Pool by about 10% recently and went from 3% Free at MAx
>Allocation to 7% Free -- so "Something" likes the new storage. Again, we
>don't have a "Problem" and I appreciate the feedback. It "Aint Broke" as
>Chris and WIllie point out so maybe move on to a more "immediate" problem.
>
> Thanks.....Ken Hynes
>
>---------------------------------------------------------------------------
------
>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
>
>
>


--------------------------------------------------------------------
mail2web - Check your email from the web at
http://mail2web.com/ .

---------------------------------------------------------------------------------
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: EDM POOL CONCERN
(in response to William Favero)
Hi Ken,

From what I've seen from other people in the list say in the past is:
the smaller the edmpool you have the more % free you need. But
conversely larger EDMPOOL's can basically ignore that number. At your
EDMPOOL size of 28000 pages, my feeling is the % free is no longer a
problem and looking at the percentage of loads makes a lot more sense.
Note: SKPT and SKCT can have their pages stolen for (DBD's, CT's, PT's
and new SKCT's and new SKPT's) which is why larger EDMPOOL's tend to be
immune from the effect of low percent free in their EDMPOOL. By the
way, last night I had a number of 0.4% free; and I haven't seen a
EDMPOOL failure in years.

Note: I also have DB2 Insight too... So you can press PF1 while the
cursor is positioned on "SKCT" and/or "SKPT" for insight's
interpretation of these numbers. I love this feature of Insight!!!

Rich Humphris
-----Original Message-----
From: DB2 Data Base Discussion List [mailto:[login to unmask email] On
Behalf Of Ken Hynes
Sent: Tuesday, December 13, 2005 10:16 AM
To: [login to unmask email]
Subject: [DB2-L] EDM POOL CONCERN


This one is a "puzzler". We have an EDMPOOL that we have been working
with.
It's Free Pages has on occasion gone as low as 2 or 3% -- no load
failures
fortunately and overall hit ration performance looks good. I've always
thought that the guideline was to keep 10-15 % free wherever possible.
We
recently adjusted the pool adding about 15% more pages and as of Tuesday
(the system cycles Sunday) were as low as 7%. I suspect Dynamic Cacheing
is
the cuplrit, but don't know of a way to "Prove" it. Nor am I aware of
any
decent means of "calculating" a good EDMPOOL size (other than the
INSTALL
CLIST).

Here's the basic Layout of the Pool:

Full Fail 0 Free......... 1890
DBD.......... 1090
SKCT......... 57
CT........... 30
SKPT......... 20424
PT........... 4509
EDM Data Space Size 30000
Full Fail 0 Free......... 0
Dynamic prep. 30000
Efficiency Requests Loads
DBD........... 29217144 353
CT............ 1095149 48
PT............ 40784287 14151
Dynamic prep.. 9875654 1765457

NOTE MAXKEEPD is set at 5000

Anyone have any thoughts? Obviously we can continue to adjust the Pool
and
add more storage, but it would be nice to have some idea as to how to
get
to an "Acceptable" value -- again I'm working with trying to get 10-15%
Free at max usage since that's what I've understood as the "recommended"
level to try to maintain. One thing I have notioced in comparison with
some
other systems I am more familiar with was the SKPT is much larger than I
am
accustomed to seeing. Again any input based on previous experience with
this would be apreciated.

Regards.....Ken Hynes

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

E-MAIL CONFIDENTIALITY NOTICE: The contents of this e-mail message and any attachments are intended solely for the
addressee(s) and may contain confidential and/or legally privileged information. If you are not the
intended recipient of this message or if this message has been addressed to you in error, please
immediately alert the sender by reply e-mail and then delete this message and any attachments. If you
are not the intended recipient, you are notified that any use, dissemination, distribution, copying, or
storage of this message or any attachment is strictly prohibited.

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

Ken Hynes

Re: EDM POOL CONCERN
(in response to Richard Humphris)
Thanks again for the responses. The "key" piece I was unsure of was
clarified by Willie's response: "However, KEEPDYNAMIC YES in conjunction
with MAXKEEPD will affect the DBM1 address space storage utilization."
Since we are on the "John Campbell Storage Diet" I am very careful about
storage utiliztion -- across the board. I appreciate the input and will
move forward from here.

.....Ken Hynes

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