1 IDUGDB2-L.ORG /home/listserv/home/db2-l June 2010, week 5
2 258 44_Re: Using .Net 2003 to connect to DB2 9 z/OS0_29_Onder.CAGATAY@HALKBANK.COM.TR31_Tue, 29 Jun 2010 16:46:13 +0300487_iso-8859-9 Hi Isaac,
i wrote an application using .Net 2003 when we were on DB2 V7 on z/OS and was able to connect to DB2 with using OLEDB Connection. After migration to V8, it's still running correctly and you can see below the established connection (My project name is Connection). But i'm not sure and got no information about the connection to V9 using .Net 2003, we will see it after migration to V9.. If you would like, i can help you with the connection using OLEDB. [...]72_A642ECB156B28142A0F5A9D9DDC01C1E15D8D939E7@GTEXCMBX01.gen.halkbank.local
261 61 25_How to shrink tablespaces12_Jim McAlpine22_jim.mcalpine@GMAIL.COM31_Tue, 29 Jun 2010 16:57:09 +0100573_ISO-8859-1 We have an application which has been rewritten to use DB2 tables instead of
temporary sequential files for the purposes of checkpoint and
restartability. The only problem with that approach is that the space used
has a very high water mark. Oracle has a shrink facility which allows
reclaimation of unused space via the "alter table" but there is no
equivalent in DB2 land. Are there any upcoming features of DB2 that would
help in these circumstances or can anyone suggest any tricks to reclaim the
space without having to do reorgs. [...]60_AANLkTim2Edt8yCdTgj7klbCryaWBiFQEMhI5Ugluschh@mail.gmail.com
323 128 58_IDUG EMEA Conference in Vienna - schedule available online10_Max Scarpa16_mscarpa@CESVE.IT31_Tue, 29 Jun 2010 18:00:09 +0200472_UTF-8 IDUG is pleased to announce the availability of the online conference
schedule for its upcoming conference in Vienna, Austria, 8-12 November
2010:
www.idug.org/emea
The schedule features speaker, company, abstract and biography information
for each session. The tool is easy-to-use and allows you to
search for sessions specific to your area of expertise.
Here is a sample of the sessions you’ll find this November in Vienna: [...]66_OF7FF0D458.337D7079-ONC1257751.0056CA69-C1257751.0057E4D2@cesve.it
452 659 44_Re: Using .Net 2003 to connect to DB2 9 z/OS12_Isaac Yassin20_yassini@BEZEQINT.NET31_Tue, 29 Jun 2010 19:09:48 +0300463_ISO-8859-1 Hi
Thanks J
We are “playing” with both .net 2003 and 2005 connecting to V9 NFM (NDS) now
and have some interesting findings .
I’ll post more when we’ll be sure about the results.
Isaac
From: IDUG DB2-L [mailto:DB2-L@IDUGDB2-L.ORG] On Behalf Of
Onder.CAGATAY@HALKBANK.COM.TR
Sent: Tuesday, June 29, 2010 4:46 PM
To: DB2-L@IDUGDB2-L.ORG
Subject: Re: [DB2-L] Using .Net 2003 to connect to DB2 9 z/OS [...]35_008401cb17a5$7f5216c0$7df64440$@net
1112 254 29_Re: How to shrink tablespaces14_Sevetson, Phil22_PSevetson@FISA.NYC.GOV31_Tue, 29 Jun 2010 12:30:27 -0400696_us-ascii If you don't have any permanent data in the DB2, there's always LOAD/REPLACE with //SYSREC DD DUMMY
-ps
________________________________
From: IDUG DB2-L [mailto:DB2-L@IDUGDB2-L.ORG] On Behalf Of Jim McAlpine
Sent: Tuesday, June 29, 2010 11:57 AM
To: DB2-L@IDUGDB2-L.ORG
Subject: [DB2-L] How to shrink tablespaces
We have an application which has been rewritten to use DB2 tables instead of temporary sequential files for the purposes of checkpoint and restartability. The only problem with that approach is that the space used has a very high water mark. Oracle has a shrink facility which allows reclaimation of unused space via the "alter table" but [...]64_1BA00CBB414DA34AA25ED82F4647CDE2033BC46A71@MAIL02.fisalan.nycnet
1367 298 29_Re: How to shrink tablespaces12_Isaac Yassin20_yassini@BEZEQINT.NET31_Tue, 29 Jun 2010 20:03:18 +0300700_us-ascii Hi,
IN DB2 9 there is the SQL "TRUNCATE TABLE creator.tbname REUSE STORAGE
IGNORE DELETE TRIGGERS IMMEDIATE "
Isaac Yassin
From: IDUG DB2-L [mailto:DB2-L@IDUGDB2-L.ORG] On Behalf Of Jim McAlpine
Sent: Tuesday, June 29, 2010 6:57 PM
To: DB2-L@IDUGDB2-L.ORG
Subject: [DB2-L] How to shrink tablespaces
We have an application which has been rewritten to use DB2 tables instead of
temporary sequential files for the purposes of checkpoint and
restartability. The only problem with that approach is that the space used
has a very high water mark. Oracle has a shrink facility which allows
reclaimation of unused space via the "alter table" [...]35_00b401cb17ac$f90fd810$eb2f8830$@net
1666 440 29_Re: How to shrink tablespaces14_Sevetson, Phil22_PSevetson@FISA.NYC.GOV31_Tue, 29 Jun 2010 15:11:42 -0400464_us-ascii Isaac,
Jim said he wanted the DASD back - TRUNCATE TABLE won't affect tablespace or index dataset size.
DROP STORAGE
Specifies that all storage that is allocated for the table is released and made available for use for the same table or any other table that resides in the table space. The scope of DROP STORAGE is always at the table space level and the deallocated space is always available for reuse by all tables in the table space. [...]64_1BA00CBB414DA34AA25ED82F4647CDE2033BC46A75@MAIL02.fisalan.nycnet
2107 81 29_Re: How to shrink tablespaces10_Dave Nance16_dav1mo@YAHOO.COM31_Tue, 29 Jun 2010 13:35:32 -0700674_iso-8859-1 Jim, What platform is this on? I think most folks are ASSuming Z/OS. David Nance 804-683-1507 ________________________________ From: Jim McAlpine To: DB2-L@IDUGDB2-L.ORG Sent: Tue, June 29, 2010 11:57:09 AM Subject: [DB2-L] How to shrink tablespaces We have an application which has been rewritten to use DB2 tables instead of temporary sequential files for the purposes of checkpoint and restartability. The only problem with that approach is that the space used has a very high water mark. Oracle has a shrink facility which allows reclaimation of unused space via the "alter table" but there is no equivalent in DB2 land. [...]43_666526.55170.qm@web57804.mail.re3.yahoo.com
2189 62 30_access method effect locksize?11_Wilson Liao16_d000660@SINA.COM31_Tue, 29 Jun 2010 21:32:48 -0400428_UTF-8 Hi Listers,
My test environment is DB2 V7 data sharing. I have a table which is
defined by segment tablespace, the locksize is SYSTEM. The structure and
content of the table can be simpled as follow:
seqno state ...
--------------------
1 A
2 B
3 C
4 A
5 B
6 C
7 A
8 B
9 C
...
The first process A update seqno 1, there should exist X LOCK on first
page. [...]48_7722999239610331.WA.d000660sina.com@www.idug.org
2252 91 34_Re: access method effect locksize?9_Mike Bell21_mbell11a1@VERIZON.NET31_Tue, 29 Jun 2010 22:18:32 -0500482_US-ASCII Basic issue is that ever since type-2 indexes there are no locks on index
pages. So a process can scan an index to find the correct row to update
without ever getting a deadlock. Once it has to read the actual row, it
will take the normal locks for the tablespace or row. As long as the target
row is on a different page, there won't be deadlock for locksize page. As
long as they identify different rows, there won't be a deadlock for locksize
row. [...]43_08F8CA350883434A83507C86202024B2@mikelaptop
2344 76 29_Re: How to shrink tablespaces12_Jim McAlpine22_jim.mcalpine@GMAIL.COM31_Wed, 30 Jun 2010 08:56:17 +0100477_ISO-8859-1 On Tue, Jun 29, 2010 at 9:35 PM, Dave Nance wrote:
> Jim,
> What platform is this on? I think most folks are ASSuming Z/OS.
>
> David Nance
> 804-683-1507
>
>
>
>
Yes, sorry the subject line should have included "DB2 for z/OS".
Jim McAlpine
_____________________________________________________________________
* IDUG Europe * Vienna, Austria * 8-12 November 2010 * http://IDUG.ORG/EU * [...]60_AANLkTim1Fovyskh3-kIbhRD1153SZM2NT8gIGOe7nVrM@mail.gmail.com
2421 30 46_translate the stmt column of syspackstmt table12_Jeremy Huang23_huangjh@SDC.ICBC.COM.CN31_Wed, 30 Jun 2010 17:30:09 +0800702_US-ASCII Dear listers,
We are in DB2 V8 for z/OS. The STMT column of SYSIBM.SYSPACKSTMT in stored
in bit data format which can't be understood by human. Is there a way to
translate them into meaningful SQL statement?
Thanks in advance!
Regards,
Jeremy Huang
_____________________________________________________________________
* IDUG Europe * Vienna, Austria * 8-12 November 2010 * http://IDUG.ORG/EU *
* Your only source for independent, unbiased, and trusted DB2 information. *
** The most DB2 technical sessions of any conference
** Access IBM experts and developers
_____________________________________________________________________ [...]69_OF6AFA93A9.93DC0C61-ON48257752.0033CBEA-48257752.00343382@icbc.com.cn
2452 39 35_Re: Tablespace in STOP PENDING Mode13_Prakash Singh30_prakash.singh@SOMERFIELD.CO.UK31_Wed, 30 Jun 2010 06:11:23 -0400627_UTF-8 Hi Mike,
Thanks a lot for clear explanation regarding the possible root cause. Really
Hi Max,
After the occurence of the problem, I got a call out. By the time I logged on, the tablespace was in STOP state and did not get anything by executing DISPLAY DB(...) Spacenam(....) USE /CLAIMERS. Generally Platinum utility executes DIS command with USE/CLAIMERS/LOCKS option when it starts and result was RW status with no locks, claiminfo or users this time. Again when there will be a contention, the utility executes the DIS command with LOCKS option so that investigation is easy for job failure. [...]62_8820897899389054.WA.prakash.singhsomerfield.co.uk@www.idug.org
2492 247 58_AW: [DB2-L] translate the stmt column of syspackstmt table35_Walter Jani=?WINDOWS-1252?Q?=DFen?=26_Walter.Janissen@ITERGO.COM31_Wed, 30 Jun 2010 13:37:49 +0200666_iso-8859-1 Hi Jeremy
You can use the CAST-function combined with the STRING-function. But these are depending on the row of a statement. The first row of each statement contains 2 lengths, after that the statement begins. After the statement-text information about all the hostvariables occur. And if you do not correctly identity the start and end position for the STRING-function you will get a -330. Unfortunatly there is no build-in-function, which translates a hexadecimal length-specification into a number, so I wrote my own litte function, which I called HEX_TO_SMALLINT. This function calls a Cobol-routine, which contains only one single MOVE. [...]43_DB2-L%201006300737593680.1C40@IDUGDB2-L.ORG
2740 140 42_AW: [DB2-L] access method effect locksize?35_Walter Jani=?WINDOWS-1252?Q?=DFen?=26_Walter.Janissen@ITERGO.COM31_Wed, 30 Jun 2010 13:42:42 +0200560_iso-8859-1 Hi
To add something, mark already pointed out. A tablespace scan starts always with the very first data page, and if the lock can't be avoided, DB2 will acquire a lock, to see, if the rows on that page qualify. If there is already an X-lock on that page, held by a different application, the scan has to wait. The same will hold for a non-matching index access. If you got a matching index access, the page, which contains the row with seqno 1, may be skipped and therefore no lock will be acquired. So locking is access path dependent. [...]43_DB2-L%201006300742523725.1C43@IDUGDB2-L.ORG
2881 309 57_Question on character representation when doing an insert13_Robert Knight28_bknight@REMOTEDBAEXPERTS.COM31_Wed, 30 Jun 2010 08:59:37 -0400454_us-ascii I have a user who executed the following piece of code:
INSERT INTO DBT1T4.EDIT_TABLE
(EDIT_TABLE_NBR, CODE_VALUE, CODE_DESC)
VALUES
(302,
'AMS 07',
'new order at myshape.com! Questions? Go');
And the results on the table have a delta instead of an exclamation point!
When I do a select against the table the DB2 representation is the following:
302 AMS 07 new order at myshape.com] Questions? Go [...]75_8902ECFB06C73440874A90F82900BA0E1B4E0F138C@osgood.rdba.remotedbaexperts.com
3191 61 50_Re: translate the stmt column of syspackstmt table13_David Simpson22_dsimpson@THEMISINC.COM31_Wed, 30 Jun 2010 09:04:02 -0400470_us-ascii This double cast works. The bit data must be first understood as
Unicode (CCSID 1208) and then re-cast as EBCDIC.
CAST(
CAST (STMT AS VARCHAR(3500) CCSID 1208)
AS VARCHAR(3500) CCSID EBCDIC
)
-----Original Message-----
From: IDUG DB2-L [mailto:DB2-L@IDUGDB2-L.ORG] On Behalf Of Jeremy Huang
Sent: Wednesday, June 30, 2010 4:30 AM
To: DB2-L@IDUGDB2-L.ORG
Subject: [DB2-L] translate the stmt column of syspackstmt table [...]67_7EC87535E7581C4C890F5BF1CF9A9374029B6F6A@THEMISSRV1.themisinc.local
3253 495 61_Re: Question on character representation when doing an insert13_Bell, Raymond20_Raymond_Bell@BMC.COM31_Wed, 30 Jun 2010 08:04:23 -0500536_us-ascii Bob,
If it's dynamically executed SQL (like SPUFI, QMF, etc) it might be down to the 3270 terminal emulation software's code page. Typically you get a warning when going into SPUFI if the subsystem and terminal's code pages are different. That might be one place to look.
Cheers,
Raymond
From: IDUG DB2-L [mailto:DB2-L@IDUGDB2-L.ORG] On Behalf Of Robert Knight
Sent: 30 June 2010 14:00
To: DB2-L@IDUGDB2-L.ORG
Subject: [DB2-L] Question on character representation when doing an insert [...]67_78B3EFE2C0C5D74E9E7777D1D669D5420FAC5289@PHXCCRPRD04.adprod.bmc.com
3749 55 50_Re: translate the stmt column of syspackstmt table13_Walters, Paul28_Paul.A.Walters@SALLIEMAE.COM31_Wed, 30 Jun 2010 09:26:25 -0400581_us-ascii Check out this technote from IBM
http://www-01.ibm.com/support/docview.wss?uid=swg21203359
-----Original Message-----
From: IDUG DB2-L [mailto:DB2-L@IDUGDB2-L.ORG] On Behalf Of Jeremy Huang
Sent: Wednesday, June 30, 2010 5:30 AM
To: DB2-L@IDUGDB2-L.ORG
Subject: [DB2-L] translate the stmt column of syspackstmt table
Dear listers,
We are in DB2 V8 for z/OS. The STMT column of SYSIBM.SYSPACKSTMT in stored
in bit data format which can't be understood by human. Is there a way to
translate them into meaningful SQL statement? [...]70_EBC218E23BF7B5468DC64F2ED8547F42770F7E008D@VREEXMBX01.us.ad.usa-ed.net
3805 128 71_Betr: [DB2-L] Question on character representation when doing an insert15_Marcel Harleman31_mam.harleman@BELASTINGDIENST.NL31_Wed, 30 Jun 2010 15:19:54 +0200448_utf-8 Bob,
we have made several packages with different CCSIDs (f.i. SPCS1140,
SPCS0500, SPCS1148 for SPUFI, TEP21140, TEP21148 etc. for DSNTEP2). We ask
the users to be aware of the codepage their session is using and use the
appropriate plan. For SPUFI there is an option under SPUFI Change Defaults:
"CHANGE PLAN NAMES" and we ask the user to select the correct plan there.
For DSNTEP2 it's in the run card of course. [...]76_OFDE502876.06795D15-ONC1257752.00491D05-C1257752.00493C00@belastingdienst.nl
3934 388 61_AW: Question on character representation when doing an insert35_Walter Jani=?WINDOWS-1252?Q?=DFen?=26_Walter.Janissen@ITERGO.COM31_Wed, 30 Jun 2010 15:40:51 +0200675_iso-8859-1 Hi
I am not an expert on that, but I suppose, you are working with different CCSIDs and the conversion is not working correctly.
May be others can give you more information.
Mit freundlichen Grüßen
Walter Janißen
ITERGO Informationstechnologie GmbH
Anwendungsentwicklung
Laufzeitarchitektur
Victoriaplatz 2
40477 Düsseldorf
mailto:walter.janissen@itergo.com
Vorsitzender des Aufsichtsrates: Jürgen Vetter
Geschäftsführung: Dr. Bettina Anders (Vorsitzende),
Ina Kirchhof, Dr. Christian Nymphius, Dr. Michael Regauer, Wolfgang Schön.
Sitz: Düsseldorf, Handelsregister: Amtsgericht Düsseldorf, HRB 37996 [...]43_DB2-L%201006300941047096.1C53@IDUGDB2-L.ORG
4323 40 54_Free Tools for DB2 Help You Improve Access Performance11_Dave Beulke19_dave@DAVEBEULKE.COM31_Wed, 30 Jun 2010 11:53:37 -04000_51_6696267950346708.WA.davedavebeulke.com@www.idug.org
4364 97 53_DB2-LUW: Block-based bufferpools and automatic tuning10_Mir, Sally15_SMir@BBANDT.COM31_Wed, 30 Jun 2010 12:20:54 -0400570_us-ascii Hello everyone,
I've crossed over from z/OS to LUW, and I'm very much a newbie, so
please forgive my ignorance. I have a simple question. When
configuring a bufferpool for block-based I/O that uses automatic tuning,
what does the threshold really mean? For example, if my bufferpool is
defined initially as 1000 pages, and I set my block size to 800, what
happens to that 800 when my bufferpool grows itself? Does it stay fixed
at 800, or does it grow at 80% of whatever my current allocation is?
Does this even make any sense? [...]62_994039AECC75DF4CAFE407FA0F32CDD00E2F3839@wil-exmb01.bbtnet.com
4462 117 58_Re: Free Tools for DB2 Help You Improve Access Performance7_Ed Long19_rdhm99a@PRODIGY.NET31_Wed, 30 Jun 2010 09:38:08 -0700473_iso-8859-1 I agree that OSC works great.
I hear that, apparently, IBM has decided to kill it.
Edward Long
--- On Wed, 6/30/10, Dave Beulke wrote:
From: Dave Beulke
Subject: [DB2-L] Free Tools for DB2 Help You Improve Access Performance
To: DB2-L@IDUGDB2-L.ORG
Date: Wednesday, June 30, 2010, 11:53 AM
A new post on my blog...
Free Tools for DB2 Help You Improve Access Performance [...]43_512436.27130.qm@web80207.mail.mud.yahoo.com
4580 170 58_Re: Free Tools for DB2 Help You Improve Access Performance10_Mark Stone17_mastone@GMAIL.COM31_Wed, 30 Jun 2010 09:57:25 -0700352_ISO-8859-1 To True, IBM had a presentation last year for Data Studio and discussed the
fact that OSC had been deprecated and no new functionality would be added,
same goes for Control Center. The free version of Data Studio is a FAR
inferior product. Surprisingly NOT!, the missing functionality is in the
Paid version of Data Studio. [...]60_AANLkTino1SyyRyor5UJamtrugpdQqOqbYqfDgF4ExEEJ@mail.gmail.com
4751 53 27_Nested loop vs. Hybrid join8_Le Hoang19_db2newbie@GMAIL.COM31_Wed, 30 Jun 2010 12:53:32 -0400490_UTF-8 Hi all,
I just ran across a problem that I didn't know the reason so I wanted to know if someone on the list could provide an explanation.
My application developer informed me that one of the query in the test environment was running extremely slow. Yet the same SQL was performing very well in production. Upon comparing the two access paths, I found that the SQL in production environment used a hybrid join method comparing to a nested loop join method in test. [...]51_8190512142449184.WA.db2newbiegmail.com@www.idug.org
4805 507 58_Re: Free Tools for DB2 Help You Improve Access Performance14_Sevetson, Phil22_PSevetson@FISA.NYC.GOV31_Wed, 30 Jun 2010 13:25:55 -0400676_us-ascii I predict that this will improve sales... of BMC DB2 Apptune and CA Insight for DB2.
________________________________
From: IDUG DB2-L [mailto:DB2-L@IDUGDB2-L.ORG] On Behalf Of Mark Stone
Sent: Wednesday, June 30, 2010 12:57 PM
To: DB2-L@IDUGDB2-L.ORG
Subject: Re: [DB2-L] Free Tools for DB2 Help You Improve Access Performance
To True, IBM had a presentation last year for Data Studio and discussed the fact that OSC had been deprecated and no new functionality would be added, same goes for Control Center. The free version of Data Studio is a FAR inferior product. Surprisingly NOT!, the missing functionality is in the Paid version [...]64_1BA00CBB414DA34AA25ED82F4647CDE2033BC46A7F@MAIL02.fisalan.nycnet
5313 89 31_Re: Nested loop vs. Hybrid join11_Mike Turner19_mike.turner@GMX.COM31_Wed, 30 Jun 2010 18:29:51 +0100322_UTF-8 Hi Hoang
Other factors beyond the statistics do influence the Optimizer. In know that
Buffer Pool size and CPU speed certainly do. It is very likely that Sort
Pool and RID Pool sizes may also do so. Some DSNZPARMs also influence the
optimizer, but I do not think they are relevant in your case. [...]44_FECBA437104E477BB2D0955E6BF07F27@ToshibaA300
5403 23 28_Ascending/Descending Indexes12_George Henke22_henke_george@YAHOO.COM31_Wed, 30 Jun 2010 15:49:59 -0400582_UTF-8 Since DB2 V8, can Descending Indexes be dropped without impacting performance because both Ascending and Descending Indexes are no longer needed for performance?
_____________________________________________________________________
* IDUG Europe * Vienna, Austria * 8-12 November 2010 * http://IDUG.ORG/EU *
* Your only source for independent, unbiased, and trusted DB2 information. *
** The most DB2 technical sessions of any conference
** Access IBM experts and developers
_____________________________________________________________________ [...]53_5342304940738109.WA.henkegeorgeyahoo.com@www.idug.org
5427 98 31_Re: Nested loop vs. Hybrid join14_Michael Kalena27_Michael.Kalena@JPMORGAN.COM31_Wed, 30 Jun 2010 17:13:16 -0400448_utf-8 Hi Hoang,
We had a very simialr case to yours. It was due to the RID Pool being dramatically smaller in Test, than Production. Once we increased the RID Pool size in Test, the query ran fine.
We are version 8. Hope this helps.
Michael Kalena | Investment Bank Technology - DB2 DBA | J.P. Morgan |
115 South Jefferson Road, Building C - NJ5-0016, Whippany, NJ 07981 | T: 973 793-2133 | Michael.Kalena@jpmorgan.com [...]75_276F3617E5150C48A5795557E7E68D200236FA1C67@EMASC201VS01.exchad.jpmchase.net
5526 473 61_Re: Question on character representation when doing an insert17_Dell'Anno, Aurora22_Aurora.Dellanno@CA.COM31_Wed, 30 Jun 2010 23:14:51 +0100419_us-ascii Bob,
if it is dynamic SQL, you should have no problem - you might see it
differently but the data is the same.
The conversion, interestingly, is not normally done by DB2 but by z/OS -
what character code is your emulator? and your DB2? and your tables?
Also, do you mind my using your example in a presentation that I have
started giving regarding "the joys" of Unicode? :-) [...]56_3D02E8610514C04F991CF832BA154C6610567964@UKSLMS11.ca.com
6000 53 32_Re: Ascending/Descending Indexes13_Willie Favero21_wfavero@ATTGLOBAL.NET31_Wed, 30 Jun 2010 17:58:53 -0500814_UTF-8 Yes, but very carefully...
Check out this blog post...
http://it.toolbox.com/blogs/db2zos/forward-or-backward-its-all-just-the-same-8394
There's some sample SQL for determining which indexes might be eligible
for delete.
And don't forget about the last used column in real time statistics.
Willie
On 6/30/2010 2:49 PM, George Henke wrote:
> Since DB2 V8, can Descending Indexes be dropped without impacting performance because both Ascending and Descending Indexes are no longer needed for performance?
>
> _____________________________________________________________________
> * IDUG Europe * Vienna, Austria * 8-12 November 2010 * http://IDUG.ORG/EU *
>
> * Your only source for independent, unbiased, and trusted DB2 information. *
> ** [...]30_4C2BCC2D.8060504@attglobal.net
6054 27 34_Re: access method effect locksize?11_Wilson Liao16_d000660@SINA.COM31_Wed, 30 Jun 2010 21:46:26 -0400604_UTF-8 Hi Mike, Walter
Thank you very much!
Best Regards
Wilson Liao
_____________________________________________________________________
* IDUG Europe * Vienna, Austria * 8-12 November 2010 * http://IDUG.ORG/EU *
* Your only source for independent, unbiased, and trusted DB2 information. *
** The most DB2 technical sessions of any conference
** Access IBM experts and developers
_____________________________________________________________________
If you need to change settings, http://www.idug.org/cgi-bin/wa?A0=DB2-L is the home of IDUG's DB2-L48_1532637740541131.WA.d000660sina.com@www.idug.org
6082 39 29_Re: How to shrink tablespaces9_Ray Lopez19_raylopez@US.IBM.COM31_Wed, 30 Jun 2010 22:11:25 -0400415_UTF-8 I agree with Phil. Since your goal is to use DB2 tables instead of temporary sequential files for the purposes of checkpoint and restartability, it sounds like you want to start with an empty table each time. Using DB2 LOAD/REPLACE with input of either a DUMMY dataset or just an empty one, will quickly and easily set you back to empty and recover the space. I have used this technique many many times. [...]51_0336945121236074.WA.raylopezus.ibm.com@www.idug.org