1 WWW.IDUGDB2-L.ORG /home/listserv/home/db2-l November 2000, week 5
2 68 32_Re: Execution of Multiple EXISTS10_teldb2kals22_teldb2kals@TELSTRA.COM31_Wed, 29 Nov 2000 16:48:49 +1000508_ISO-8859-1 Hi Suresh,
Yes, DB2 is smart enough to stop processing if the first EXISTS is true, provided they are correlated subqueries. My understanding is that, if the subqueries are noncorrelated ones, then DB2 executes both the subqueries to completion, and only then executes the OR clause.
Cheers, Kals
-----Original Message----- From: Suresh Sane [SMTP:data_arch@HOTMAIL.COM] Sent: Wednesday, November 29, 2000 2:17 AM To: DB2-L@RYCI.COM Subject: Execution of Multiple EXISTS [...]
71 41 35_Does an elephant fly? No? Prove it.19_Alekos Papadopoulos13_apapad@NBG.GR31_Wed, 29 Nov 2000 09:24:13 +0200631_ISO-8859-7 Hello List,
We use DB2 V5 for OS390 V2.5. We plan to add some columns in some tables (by using ALTER). Several PL/I programs are accessing these tables by using 'SELECT * INTO :host_structure WHERE uniqueindex=something'. We use DCLGEN to create the host_structure. In page 2-6 of APSG the manual suggests that using 'SELECT *' is not recommended in static SQL, because it might cause reference to columns for which we have not defined receiving host variables. Based on that, we proposed to the management that the application programmers should replace 'SELECT *' by 'SELECT column1, column2 ...). At this [...]
113 56 36_Re: accessing DB2 V6.1 from Db2 UDB.0_24_markus.schaub@SYSTOR.COM31_Wed, 29 Nov 2000 08:14:21 +0100365_us-ascii to access tables on a os/390 you have to use the drda protocol. you will get the support for this protocol through the db2 connect product or you can install a enterprice edtion of udb which includes the connect code.
if you want to work with the control center with db2 on os/390 you have to install some stored procedures on the os/390 side. [...]
170 83 12_Re: Packages16_Aurora Dell'Anno26_aurora.dellanno@SISTINF.IT31_Wed, 29 Nov 2000 00:08:37 +0100473_us-ascii Ronald,
plug for somebody else's product here - I believe ENDEVOR might do a lot of your dirty work here, and I have worked with it in the past with very satisfactory results.
HTH.
Aurora Emanuela Dell'Anno Systems Programmer Sistemi Informativi S.p.A. e-mail: aurora.dellanno@sistinf.it tel. +39 335 7429486
no trees were killed in the sending of this message. however - a large number of electrons were seriously inconvenienced. [...]
254 102 32_Re: Execution of Multiple EXISTS16_S, Sanjeev (CTS)25_SSanjeev@CAL.CTS-CORP.COM31_Wed, 29 Nov 2000 13:02:02 +0530522_- This is what i was trying to recall from the morning i.e Co-related .Kals has mentioned logically that DB2 will have to evaluate both the EXISTS when non-corelated subquery is used. As far as performance point of view is concern, if we put the efficiently accessed table first then query can be really faster.When we are using EXISTS then the access path more or less is dependent on how the outer table is accessed.The number of touches in the inner table is always determined by the outer tables qualifying rows. [...]
357 60 15_Re: CURRENTDATA16_S, Sanjeev (CTS)25_SSanjeev@CAL.CTS-CORP.COM31_Wed, 29 Nov 2000 13:17:55 +0530446_- Hi Murali, There are corelation between CURRENT DATA and Isolation levels but it is not the way that CURRENT DATA(NO) will not take the locks as UR. CURRENT DATA (NO) tries to use the lock avoidance technique but it is not neccessary that it will always get success but UR reads the uncommited data. In my opinion, Answer to your question is ofcourse YES but it is not neccessarily because of CURRENT DATA (NO) is able to avoid the locks. [...]
418 100 12_Re: Packages14_Ruediger Kurtz28_ruediger.kurtz@HUK-COBURG.DE31_Wed, 29 Nov 2000 09:05:06 +0100595_us-ascii Ronald,
we're currently allowing versioning but are only using 3 versions, namely 2, 4 and 8. As we have 3 load-libraries concatenated (search me for the reason for this, as this was implemented long before I joined this company, I think it was mainly for fall-back situations when you just delete the latest version and the last-but-one will become effective), we had to adopt this 3-library-approach to our dbrm-libraries as well. In the course of time our load- and dbrm-members are being moved from library 1 to no. 2 and finally to 3, so that in the beginning we might [...]
519 153 38_Re: stop/start database in batch mode.0_18_mebert@AMADEUS.NET31_Wed, 29 Nov 2000 09:51:31 +0100508_us-ascii This is still on DB2 V5, 9911 level or thereabouts, with a couple of PTFs. I do remember that some time last year I asked on this list how to cancel a batch job trying to do a -STOP; the only solution then was to FORCE the job. After someone on the list said that this might also force you to re-IPL, I got sort of hesitant about using this. A couple of days ago I tried to -STOP a TS that was in use... and lo and behold, the cancel worked. We certainly didn't apply any maintenance for this. [...]
673 80 39_Re: Does an elephant fly? No? Prove it.16_S, Sanjeev (CTS)25_SSanjeev@CAL.CTS-CORP.COM31_Wed, 29 Nov 2000 03:58:42 -0500566_- Hi Alekos, Most important thing is maintenance of the programs are getting a lot of difficult by using SELECT * .....While doing the impact analysis for a particular change you are finding problems and if the columns would have been coded instead of * then impact analysis would have been more easier. I am giving one example which i think may suits to your condition. Suppose SELECT * INTO : DCLGEN VARS FORM T1 is writting in the program(A) and currently T1 contains some 10 columns and your program is running nicely.Next you have added a new column to T1 [...]
754 23 28_Re: New IBM DB2 tools -Reply0_29_paul.packham@POSTOFFICE.CO.UK31_Wed, 29 Nov 2000 09:46:47 +0000352_us-ascii James,
We are doing an evaluation of the Log Analysis Tool for OS/390. Administration Tool for OS/390 V2R1, DB2 SQL Performance Analyser for OS/390 V1R1M0 & DB2 Query Monitor for OS/390.
Its a little early to comment on the functionality & quality of the products. When we complete the work I'll post something to the list. [...]
778 35 34_Typrun=SCAN for SQL statements...?10_Max Scarpa16_mscarpa@CESVE.IT31_Wed, 29 Nov 2000 11:08:37 +0100595_us-ascii Hi folks.
We are building a new procedure (using REXX/ISPF skeletons/Control-M, in a perfect autarchy) to automate the creation of new DB2 objects, from a test environment to the production environment.
I'm wondering if there's a simple way to 'test' the SQL statements, i.e. to do something similar to TYPRUN=SCAN in jcl, a check without execution. We can use the EXPLAIN statement, but I'd like to know if there's another way to do this check WITHOUT a wonderful 3rd party toolz (OK if CDB can send a totally free product to do the check be sure I'll wear all they [...]
814 77 38_AW: Typrun=SCAN for SQL statements...?12_Harvey,David19_David.Harvey@TUI.DE31_Wed, 29 Nov 2000 11:25:46 +0100356_iso-8859-1 Max, Presumably in your (sophisticated) dialogue/dialog you determine Source+Target SSID and DB dynamically i.e. NOT hard-coded. Another option to code into the rexx would be "Rollback Y/N" and then you could get what you want very cheaply. The rexx code could simply add the Rollback before the end of each LUW if that option were chosen. [...]
892 229 8_CPU-Cost14_Ruediger Kurtz28_ruediger.kurtz@HUK-COBURG.DE31_Wed, 29 Nov 2000 12:28:18 +0100320_us-ascii Hi folks,
one of our application designers approached me with a question I have no answer to:
He has a join on three tables and the used CPU-Time is considerably higher compared to selecting data from the tables one after the other (2 minutes vs 9 secs);
Here's the SQL with the join: [...]
1122 24 49_Typrun=SCAN for SQL statements...? errata corrige10_Max Scarpa16_mscarpa@CESVE.IT31_Wed, 29 Nov 2000 12:31:10 +0100333_us-ascii In my previous mail I wrote :
"I'm wondering if there's a simple way to 'test' the SQL statements, i.e. to do something similar to TYPRUN=SCAN in jcl, a check without execution. "
Of course we need to test SQL AND DDL so we need something to 'check' DDL, too.
Regards
Max Scarpa DB2 sysprog [...]
1147 97 39_Re: Does an elephant fly? No? Prove it.16_S, Sanjeev (CTS)25_SSanjeev@CAL.CTS-CORP.COM31_Wed, 29 Nov 2000 15:53:04 +0530580_- Correction to my previous mail : Plan A will not be invalidated unless until the new column added is of DATE,TIME or TIMESTAMP data type but if we run program A without changing its DCLGEN then particular SQL will return -ve sqlcode
HTH
Regards Sanjeev
> -----Original Message----- > From: S, Sanjeev (CTS) > Sent: Wednesday, November 29, 2000 2:27 PM > To: 'DB2 Data Base Discussion List' > Subject: RE: Does an elephant fly? No? Prove it. > > Hi Alekos, > Most important thing is maintenance of the programs are getting a > lot of difficult by using [...]
1245 122 39_Re: Does an elephant fly? No? Prove it.25_Sinha, Ranjay Kumar (CTS)24_SRanjay@CAL.CTS-CORP.COM31_Wed, 29 Nov 2000 15:59:59 +0530508_iso-8859-1 Hi Sanjeev, I am not comfortable with the following lines from ur reply. "As soon as the Alter is issued DB2 will invalidate the plan A and it will attempt a rebind in the ....." As far as I know " when ever a table is altered for a column addition the plan and packages are not invalided untill unless the new column are TIME,DATE or TIMESATMP". I really do not know why the plan is marked "invalid" in case of TIME,DATE or TIMESTAMP. Hope to get more opinion from the list TIA Ranjay sinha [...]
1368 24 17_'Data Migration '0_18_dsi.listsrv@DB.COM31_Wed, 29 Nov 2000 17:12:13 +0530600_us-ascii Hi,
We wanted to transfer large volume of data from UDB5.2 on NT to UDB 6.x on MainFrame (OS/390). Is there any easy way to do this ?? ( ie. using Export/Import utility) ??
Regards Siva
============================================================================================= DISCLAIMER: The opinion expressed above is that of the author and should not be understood or in-terpreted as equivalent to the opinion of Deutsche Bank AG or any of its subsidiary companies ============================================================================================= [...]
1393 63 21_Re: 'Data Migration '16_S, Sanjeev (CTS)25_SSanjeev@CAL.CTS-CORP.COM31_Wed, 29 Nov 2000 17:25:29 +0530527_- Hi Siva, If it is a one time job, i think it is better to use Unload + File transfer and load but some good amount of effort will be required to create the punch card on mainframe. Will DB2 V6 control center for OS/390 (ofcourse on NT) will be able to do that ? If large number of tables are involved and creation of punch will be a tedious task then it is better to go for replication when your network traffic is low i.e lesser workload on network. Export/Import will again require some more manual efforts if control [...]
1457 68 21_Re: 'Data Migration '10_Alan Johns26_AlanJ@PRINCETONSOFTECH.COM31_Wed, 29 Nov 2000 07:30:22 -0500322_ISO-8859-1 Hi Siva. I don't know whether you would want to look at 3rd party tools for this, but there is a tool called Move for Servers, from Princeton Softech. It is a relational data extraction and insertion tool which has the capability to move data between different DBMS's, including DB2 UDB and DB2 mainframe. [...]
1526 192 86_=?ISO-8859-7?Q?=C1=D0=3A_Does_an_elephant_fly=3F_No=3F_Prove_i?= =?ISO-8859-7?Q?t=2E?=19_Alekos Papadopoulos13_apapad@NBG.GR31_Wed, 29 Nov 2000 14:43:21 +0200541_ISO-8859-7 To be more specific: I created a table named TESTTB with just 2 columns (CODE DEC(3) NOT NULL and DESCR CHAR(10) NOT NULL) I added a few rows to it. I generated a PLI record from DCLGEN and I included it in a program that had just does an EXEC SQL SELECT * INTO :DCLTESTTB (which is the 01 level of the structure) FROM TESTTB WHERE CODE =1; and then writes the two columns in the outlist. I then altered the table with the following statement ALTER TABLE TESTTB ADD XTCOL CHAR(5) NOT NULL WITH DEFAULT 'ZZZZZ'; The ALTER was [...]
1719 31 39_Re: Does an elephant fly? No? Prove it.13_Seibert, Dave26_Dave_Seibert@COMPUWARE.COM31_Wed, 29 Nov 2000 07:38:34 -0500454_iso-8859-1 Hi Alekos,
If you select more columns than you provide host variables, you don't get a negative SQLCode (personally I think you should).
Rather you get SQLWARN3 turned on.
Is your test program evaluating SQLWARNs?
I strongly recommend that programs should evaluate SQLWARNings. That doesn't mean that there's always reason for fatal error or abend. Look over the description of SQLCA to determine if you care. [...]
1751 55 38_Re: Typrun=SCAN for SQL statements...?14_Grainger, Phil20_Phil.Grainger@CA.COM31_Wed, 29 Nov 2000 12:48:38 -0000407_iso-8859-1 Provided a valid PLAN_TABLE exists, an EXPLAIN of the statement should indicate whether it is syntactically valid or not.
Depending on how complex your environment is, a PREPARE also be a good choice!
Phil Grainger DB2 Product Manager Computer Associates
PS I'd be glad to sell you a tool to do this, but there doesn't seem to be much point when you can do it for free! [...]
1807 55 53_Re: Typrun=SCAN for SQL statements...? errata corrige14_Grainger, Phil20_Phil.Grainger@CA.COM31_Wed, 29 Nov 2000 12:51:56 -0000395_iso-8859-1 The problem with checking DDL as opposed to SQL is that there is a difference between a SYNTAX check and a SEMANTICS check.
eg
Consider two SQL statements:
CREATE DATABASE FRED; GRANT DBADM ON DATABASE FRED TO BERT;
These will both syntax check OK, but unless the first statement is executed the second one is invalid but STILL syntactically correct. [...]
1863 15 12_Re: Packages15_Walter Janissen27_walter.janissen@VICTORIA.DE31_Wed, 29 Nov 2000 06:53:23 -0600508_- Hi Scott
We keep it very simple. We only use one collection in production for all packages, it doesn't matter whether main or subroutine and we bind all plans with PKLIST(Collection.*). We use this concept for years and we haven't got any problem with that.
================================================ To change your subscription options or to cancel your subscription visit the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list can be reached at DB2-L-REQUEST@RYCI.COM.
1879 39 12_Re: Packages14_Scott Trometer22_scott.trometer@RCI.COM31_Wed, 29 Nov 2000 08:03:54 -0500388_iso-8859-1 I agree. I have seen this application before, as well. I agree that it is very simplistic, but I also know that many folks would take issue with it being a performance risk...I believe you said the magic words, however, when you said that you haven't experienced any problems.
I think, especially with DB2, the adage, if it ain't broke don't fix it REALLY applies. [...]
1919 64 38_Re: Typrun=SCAN for SQL statements...?24_Peter_Schwarcz (Bigpond)26_Peter_Schwarcz@BIGPOND.COM31_Thu, 30 Nov 2000 00:13:00 +1000358_us-ascii Max,
I think that the equivalent of a TYPRUN=SCAN would be to run a PREPARE on an SQL statement.
You could modify one of the IBM supplied dynamic SQL programs (such as DSNTIAD) to just do a prepare and set a return code as appropriate.
Once you are in the code you could also return a cost estimate for each SQL statement. [...]
1984 26 33_WLM Stored Procedure VARY command18_Kimball, William P19_KimballWP@AETNA.COM31_Wed, 29 Nov 2000 08:19:53 -0500579_iso-8859-1 We currently have a few DB2 managed stored procedures. SYSOPR was granted to the DBA group so that the DBAs could stop/start the stored procedures. Now that we have a WLM managed SP address space, we'll need to start using the VARY WLM,...... command to start the stored procedures. We've been told by the MVS system programmers that they'll never open this up to us (the DBAs), never mind the developer who might be creating these. What has anyone done to get around this MVS command requirement that system programmers may not want to let out of their control? [...]
2011 61 38_DFSORT Problems during Parallel Reorgs13_Lockwood Lyon19_Lockwool@MEIJER.COM31_Tue, 28 Nov 2000 16:32:43 -0500478_ISO-8859-7 Anyone else have this problem? We are running multiple (sometimes 6-8 at a time) Tablespace Reorgs on weekends. Most of the tablespaces are partitioned; most have Compress Yes; some may have more than one index. We prefer to run multiple Reorg utilities in the same step. In most cases, we allow DFSORT to dynamically allocate SortWork datasets. We use the Reorg "opportunity" to do in-line Image Copy and RunStats. We are DB2 V6 in a 4-member parallel sysplex. [...]
2073 59 41_Repost: Parallel Reorgs and DFSORT Issues13_Lockwood Lyon19_Lockwool@MEIJER.COM31_Wed, 29 Nov 2000 08:32:22 -0500581_ISO-8859-7 My apologies for the re-post. I didn't see the first one come through on the list.
- - -
Anyone else have this problem? We are running multiple (sometimes 6-8 at a time) Tablespace Reorgs on weekends. Most of the tablespaces are partitioned; most have Compress Yes; some may have more than one index. We prefer to run multiple Reorg utilities in the same step. In most cases, we allow DFSORT to dynamically allocate SortWork datasets. We use the Reorg "opportunity" to do in-line Image Copy and RunStats. We are DB2 V6 in a 4-member parallel sysplex. [...]
2133 106 42_Re: DFSORT Problems during Parallel Reorgs13_Lynne Flatley17_LFlatley@NEFN.COM31_Wed, 29 Nov 2000 08:46:12 -0500294_iso-8859-7 Could you get StopX37 turned off for your DB2 utilities? I always found it ironic that at my last place of employment, StopX37 consistently caused B37 abends on DB2 reorgs. We were not parallel sysplex and we were V4 at the time, but the end result was the same...B37 abends... [...]
2240 21 13_abbreviations13_Jimmy Willett24_Jimmy.Willett@NCMAIL.NET31_Wed, 29 Nov 2000 08:51:15 -0500353_us-ascii Does IBM have a manual or list of all the abbreviations used in DB2, CICS, and IMS??? It seems like every article you read uses abbreviations without giving the full name elsewhere. I spend as much time looking up what they mean as reading the article. Kind of confusing to read about something and not know what the meanings are.. thanks [...]
2262 30 90_Re: =?ISO-8859-7?Q?=C1=D0=3A_Does_an_elephant_fly=3F_No=3F_Prove_i?= =?ISO-8859-7?Q?t=2E?=13_Mohammad Khan20_mkkhan88@HOTMAIL.COM31_Wed, 29 Nov 2000 07:50:22 -0600625_- On Wed, 29 Nov 2000 14:43:21 +0200, Alekos Papadopoulos wrote:
>The ALTER was successfull. >Without runing DCLGEN, nor recompiling nor even rebinding manually, I >executed the PLI program again. Believing or not, the execution was >successfull, and the old columns were displayed in the outlist, without any >error.
You won't get any problem because the "SELECT *" gets frozen (i.e. actual column names are substituted ) when you bind the package and any changes do NOT have any effect on this. The problem will arise when you bind ( rebind ) a package BEFORE creating the new DECLGEN as [...]
2293 20 13_BMC Copy Plus13_Johnson, Bill18_bjo23@ALLSTATE.COM31_Wed, 29 Nov 2000 07:52:47 -0600451_- All, We have a situation in which we perform an incremental copy using BMC's copy plus with shrlevel change. Copy plus therefore issues a quiesce at the beginning which causes contention with a thread that doesn't commit very frequently. Changing the thread is not an option. Is there a way we can circumvent the quiesce issued by copy plus and if so is that wise? We are DB2 V6, Copy Plus V5.7. Thanks, Bill Johnson DBA Allstate Insurance Co. [...]
2314 59 17_Re: abbreviations14_Davage, Marcus29_Marcus.Davage@LLOYDSTSB.CO.UK31_Wed, 29 Nov 2000 13:54:56 -0000643_iso-8859-1 Take a look at http://www-3.ibm.com/ibm/terminology/ - that should start you off on your TLA quest!
Marcus
-----Original Message----- From: Jimmy Willett [mailto:Jimmy.Willett@NCMAIL.NET] Sent: 29 November 2000 13:51 To: DB2-L@RYCI.COM Subject: abbreviations
Does IBM have a manual or list of all the abbreviations used in DB2, CICS, and IMS??? It seems like every article you read uses abbreviations without giving the full name elsewhere. I spend as much time looking up what they mean as reading the article. Kind of confusing to read about something and not know what the meanings are.. thanks [...]
2374 29 17_Re: abbreviations13_Seibert, Dave26_Dave_Seibert@COMPUWARE.COM31_Wed, 29 Nov 2000 09:01:37 -0500524_iso-8859-1 There are 2 places I look for the resolution of unknown acronyms:
http://www.mtnds.com/af/
and
http://www2.ibmlink.ibm.com/cgi-bin/master?xh=VOOWT2JyWwXK*n0USenGnI9332&req uest=text&parms=mainPage.txt&xhi=&xfr=F
The first is a generic acronym finder and contains many computer acronyms, but not just computer-related and certainly not just IBM.
The second is a dictionary service within IBMLink. I believe it's publicly available -- ie. doesn't require an IBMLink account. [...]
2404 46 37_Re: WLM Stored Procedure VARY command11_David Nance16_DWNance@FHSC.COM31_Wed, 29 Nov 2000 09:00:00 -0500603_US-ASCII At my previous assignment, there was a job on hold that would stop each of the spas and we had access to release the job. Also, the job completion triggered itself to start up and go to hold status again.
Dave Nance First Health Services, Corp. (804)527-6841
>>> KimballWP@AETNA.COM 11/29/00 08:19AM >>> We currently have a few DB2 managed stored procedures. SYSOPR was granted to the DBA group so that the DBAs could stop/start the stored procedures. Now that we have a WLM managed SP address space, we'll need to start using the VARY WLM,...... command to start the stored [...]
2451 36 17_Re: abbreviations15_Lankester, Andy22_Andy_Lankester@BMC.COM31_Wed, 29 Nov 2000 08:07:24 -0600652_iso-8859-1 Check out www.gurunet.com for a windows based look-up tool - I find it veru useful - just Alt-Click a word (or acronym) ad it will look it up for you.
Andy
-----Original Message----- From: Jimmy Willett [mailto:Jimmy.Willett@NCMAIL.NET] Sent: 29 November 2000 13:51 To: DB2-L@RYCI.COM Subject: abbreviations
Does IBM have a manual or list of all the abbreviations used in DB2, CICS, and IMS??? It seems like every article you read uses abbreviations without giving the full name elsewhere. I spend as much time looking up what they mean as reading the article. Kind of confusing to read about something and [...]
2488 44 53_Re: Typrun=SCAN for SQL statements...? errata corrige10_Max Scarpa16_mscarpa@CESVE.IT31_Wed, 29 Nov 2000 15:08:19 +0100336_us-ascii Hi Phil & others listers and thanks for your reply.
I agree with Phil, SQL is easier to check than DDL but :
1) We have a standard naming, so object name must obey to some restricted rules (so the life is easiest for SMS people)
2) There's only 1 creator which owns all objects. There are few grants. [...]
2533 129 75_Re: =?Windows-1253?b?wdA6IERvZXMgYW4gZWxlcGhhbnQgZmx5PyBObz8gUHJvdmUgaXQu?=9_Colin Fay13_cfay2@CSC.COM31_Wed, 29 Nov 2000 09:28:29 -0500521_Windows-1253 Alekos,
I would like to see you Reorg the table you altered , prior to running the program. Remember an ALTER does not physically change the row layout on the page. You may have been lucky in the DB2 is passing you the old record layout (at the original length). Perhaps you may have memory overlay after you reorg the table, physically change the row length and retrieve it iinto you application. If not I sduspect that PL1 handles working storage allocation fundamentally different from COBOL. [...]
2663 166 39_Re: Does an elephant fly? No? Prove it.13_Vaughan, Mike26_Vaughan.Mike@PRINCIPAL.COM31_Wed, 29 Nov 2000 08:26:03 -0600602_iso-8859-7 This suprised me somewhat. I would have expected this to give an error instead of a warning (SQLWARN3 from Dave Seibert's message). I would still try to avoid coding "SELECT *". I don't know if it would be worth a "long and costly project" to change all of your existing programs, but would recommend avoiding it from now on (possibly changing the existing programs along with other maintenance). Consider another example -- your table initially has (ID INTEGER, START_DT DATE, END_DT DATE) and instead of altering to add a column, you drop/re-create the table so it now looks like (ID [...]
2830 54 17_Re: BMC Copy Plus15_Lankester, Andy22_Andy_Lankester@BMC.COM31_Wed, 29 Nov 2000 08:32:32 -0600372_iso-8859-1 Bill,
SLCHGQSC=YES (default installation option):
When using SHRLEVEL CHANGE and data sharing with incremental copies, specifies whether COPY PLUS is to issue a QUIESCE when no unique LSRN/RBA is found to register the copy. The default is YES. If SLSCHGQSC=NO and no unique registration is found, COPY PLUS stops and issues an error message. [...]
2885 70 15_Platinum Course19_Lisa Westcott-Dryer32_lisa_westcott_dryer@FOREMOST.COM31_Wed, 29 Nov 2000 09:54:33 -0500388_iso-8859-1 A corporate decision has been made to move from BMC to Platinum products. Definately not my decision.
Can anyone give me a handle on a couple of good ways they've learned Platinum products? Class, CBT etc. I've already started pouring through the manuals. I need specifically: Detector RC/Migrator
thanks for the help, this really is a good group of people, [...]
2956 102 19_Re: Platinum Course11_Road Runner22_hhardy1@HOUSTON.RR.COM31_Wed, 29 Nov 2000 09:35:40 -0600611_iso-8859-1 I'll forward this to someone who can assist you.... -----Original Message----- From: Lisa Westcott-Dryer Newsgroups: bit.listserv.db2-l To: DB2-L@RYCI.COM Date: Wednesday, November 29, 2000 8:59 AM Subject: Platinum Course
A corporate decision has been made to move from BMC to Platinum products. Definately not my decision.
Can anyone give me a handle on a couple of good ways they've learned Platinum products? Class, CBT etc. I've already started pouring through the manuals. I need specifically: Detector RC/Migrator [...]
3059 96 19_Re: Platinum Course0_22_BILL_GALLAGHER@PHL.COM31_Wed, 29 Nov 2000 10:16:30 -0500569_iso-8859-1 Lisa,
First of all, let me offer my condolences to you.
I would suggest that you or your management put some pressure on CA to send some onsite support to you to help you through the transition, and to give you a crash course on how to use the products. We went through a similar transition recently, and when that process started over 1 1/2 years ago, when it was still Platinum, they offered to send whatever onsite support we felt we needed to make it through the transition. However, once the deal was completed, CA had taken over and [...]
3156 122 27_Re: Package Version Cleanup11_Hayden, Lee18_Lee_Hayden@CSX.COM31_Wed, 29 Nov 2000 10:26:22 -0500583_iso-8859-1 Here is a little DB2/REXX that will generate free for all older than n number of versions. It works for auto or manual versions.
//* //* REXX/DB2 EXAMPLE OF CONTROL CARD GENERATION //* //* 1. CREATE REXX PROC INLINE (MAKES DEBUGGING SIMPLER) //* 2. EXECUTE //* //CRE#REXX EXEC PGM=IEBUPDTE, // PARM=NEW //SYSPRINT DD SYSOUT=* //SYSUT2 DD DSN=&&SYSPROC, // DISP=(NEW,PASS), // UNIT=SYSDA, // SPACE=(TRK,(5,1,25)), // DCB=(LRECL=80,RECFM=FB,BLKSIZE=0) //SYSIN DD DATA, // DLM='\\' ./ ADD NAME=REXXDB2 /* REXX NAME: REXXDB2 FUNCTION: CALL IBM REXX DB2 SUPPORT */ [...]
3279 202 39_Re: Does an elephant fly? No? Prove it.16_Provenzola, Tony24_Tony.Provenzola@NIKE.COM31_Wed, 29 Nov 2000 07:28:20 -0800371_iso-8859-7 1. Isn't it more efficient to retrieve just what you need rather than to retrieve the entire row?
2. It's better documentation to retrieve specific columns. If I had to make program changes to support a modified column on a table, I would rather know the dozen program that use that column rather than the hundreds of programs that use that table. [...]
3482 29 19_Re: Platinum Course10_Max Scarpa16_mscarpa@CESVE.IT31_Wed, 29 Nov 2000 16:26:41 +0100345_us-ascii If Detector and Migrator are the only CA/Platinum products you are going to change I think you will not have problems.
Detector is a nice product and Migrator works well and in my opinion is a little bit better than Change Manager (but more complicated)
BUT if you have some other products, IN PRIMIS utilities then: [...]
3512 96 19_Re: Platinum Course17_Fiorese Giampaolo29_Fiorese.Giampaolo@GENERALI.IT31_Wed, 29 Nov 2000 16:22:26 +0100667_iso-8859-1 What a trusted company!!!
-----Original Message----- From: BILL_GALLAGHER@PHL.COM [mailto:BILL_GALLAGHER@PHL.COM] Sent: Wednesday, November 29, 2000 4:17 PM To: DB2-L@RYCI.COM Subject: Re: Platinum Course
Lisa,
First of all, let me offer my condolences to you.
I would suggest that you or your management put some pressure on CA to send some onsite support to you to help you through the transition, and to give you a crash course on how to use the products. We went through a similar transition recently, and when that process started over 1 1/2 years ago, when it was still Platinum, they offered to send whatever onsite [...]
3609 43 22_Platinum Course -Reply12_John Bucaria22_JBUCARIA@MAIL2.PCH.COM31_Wed, 29 Nov 2000 10:28:38 -0500585_US-ASCII Lisa,
We did this transition before CA acquired Platinum. Included in the purchase of the tools was about 5 to 7 days of training at our site by qualified Platinum instructors. I recommend this approach mainly because I don't feel that the manuals are particularly user friendly. Better to get a sound foundation in the tools and then use the manuals to obtain the finer points. Also, if you're going to take this approach (classroom instruction), it's a good idea to come up with a list of what specific functions you need to have covered so that your time isn't [...]
3653 25 26_FW: IBM Interpreter Needed14_Scott Trometer22_scott.trometer@RCI.COM31_Wed, 29 Nov 2000 10:49:07 -0500507_iso-8859-1 > List, > > Can anyone explain the following as taken from the V5 SQL reference under > Alter table, Add Column : > > Adding a column to table T only changes the description of T. If the > catalog description of T is used to create a > table T' and a facility such as DSN1COPY is used to effectively copy T > into T', queries that refer to the added > column in T' will fail because the data does not match its description. To > avoid this problem, run the REORG > utility against the table [...]
3679 23 19_Re: Platinum Course0_29_paul.packham@POSTOFFICE.CO.UK31_Wed, 29 Nov 2000 15:51:08 +0000474_us-ascii Lisa,
I attended a presentation given by Phil Granger from CA at European IDUG. (CA now own Platinum). The session was about the support CA would give if they were called upon to replace other ISV's Software. Basically they were offering a package which included impact analysis on changing the software, project management/technical support of the actual change over & education. I would speak to your CA rep. to see if they can provide you with this. [...]
3703 52 12_Re: Packages13_Davis, Ronald17_DavisRo@CONED.COM31_Wed, 29 Nov 2000 11:09:37 -0500443_iso-8859-1 Thanks everyone for your feedback regarding my 'package' decision. A special thanks to Lee Hayden for the rexx exec. This is what I have so far:
1. We will have 2 plans for each application. An online and a Batch. 2. Versioning would be nice to have but I can't justify the maintenance. Our production move procedures already provide a level of fall-back. 3. I guess we're keeping it as simple as possible for our shop. [...]
3756 23 11_DDF Failure15_Daniel Sullivan28_daniel.sullivan@ZURICHUS.COM31_Wed, 29 Nov 2000 10:19:24 -0600561_us-ascii A co-worker is trying to go from one DB2 on one lpar to another DB2 on another lpar using Websphere JAVA and DDF. She is receiving the following error:
DB2SQLJConnection error in native method: constructor SQLSTATE=57011 and SQLCODE=-904
java.sql.SQLException: DB2SQLJConnection error in native method: constructor SQLSTATE=57011 and SQLCODE=-904
My question where can I find out what the unavailable resource is? I have looked in the calling DB2 and the receiving DB2 but do not see anything, any ideas would be appreciated. [...]
3780 58 26_Re: IBM Interpreter Needed16_Provenzola, Tony24_Tony.Provenzola@NIKE.COM31_Wed, 29 Nov 2000 08:29:12 -0800455_iso-8859-1 If I alter table T to add a column, it only updates the catalog, the underlying data hasn't changed. Each row of the physical data is the same length as it was before.
If I create another table, T', modeled after T (including the new column), then try to load it from T using DSN1COPY, which doesn't use the catalog, the data will not match. For instance, if I added column 11 to T, I am loading 10 columns to a table expecting 11. [...]
3839 69 26_Re: IBM Interpreter Needed17_Kornblum, Kenneth24_Kenneth_Kornblum@BMC.COM31_Wed, 29 Nov 2000 10:31:21 -0600500_iso-8859-1 Scott,
Since no one has replied yet to your query I'll give it a whack.
The manual reference you quoted is trying to tell you that the SQL
ALTER TABLE my.table ADD column-definition
changes the DB2 catalog table definition but does NOT alter the data in the tablespace. New rows inserted after ALTER would contain the new column but existing rows are not modified by ALTER. In the TS affected by the ALTER, DB2 knows how to deal with this "inconsistency." [...]
3909 43 15_Re: DDF Failure14_Scott Trometer22_scott.trometer@RCI.COM31_Wed, 29 Nov 2000 11:31:48 -0500439_iso-8859-1 Typically, the -904 sqlcode is accompanied by: a) the resource type, b) resource name c) reason code
If you can determine this information, you can refer to the DB2 messages and codes manual to determine the type and reason code.
-----Original Message----- From: Daniel Sullivan [mailto:daniel.sullivan@ZURICHUS.COM] Sent: Wednesday, November 29, 2000 11:19 AM To: DB2-L@RYCI.COM Subject: DDF Failure [...]
3953 22 23_A simple ZPARM Question14_Richard Arnold24_Richard.Arnold@CHASE.COM31_Wed, 29 Nov 2000 10:16:45 -0600538_us-ascii All,
From the manuals and it appears that size of the EDM pool cannot be adjusted "on the fly", and that it requires the ZPARM to changed and re-assembled, and the subsystem bounced.
Can someone verify that I have this right, or correct my if not?
Thanks in advance.
================================================ To change your subscription options or to cancel your subscription visit the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list can be reached at DB2-L-REQUEST@RYCI.COM.
3976 46 27_Re: A simple ZPARM Question0_25_Ed_Vetock@NAVYFEDERAL.ORG31_Wed, 29 Nov 2000 11:41:11 -0500555_us-ascii You're right.
Richard Arnold @RYCI.COM> on 11/29/2000 11:16:45 AM
Please respond to DB2 Data Base Discussion List
Sent by: DB2 Data Base Discussion List
To: DB2-L@RYCI.COM cc:
Subject: A simple ZPARM Question
All,
From the manuals and it appears that size of the EDM pool cannot be adjusted "on the fly", and that it requires the ZPARM to changed and re-assembled, and the subsystem bounced. [...]
4023 49 27_Re: A simple ZPARM Question12_Gerald Hodge20_ghodge@ATTGLOBAL.NET31_Wed, 29 Nov 2000 10:47:14 -0600539_us-ascii Richard:
Version 7 of DB2 is expected to offer the ability to change the EDM* keywords in the DSN6SPRM macro. These changes are not necessarily immediate. Before then, I believe, BMC has the only tool that will adjust the EDM Pool Size.
Gerald Hodge HLS Technologies, Inc. 888-494-9019
-----Original Message----- From: DB2 Data Base Discussion List [mailto:DB2-L@RYCI.COM]On Behalf Of Richard Arnold Sent: Wednesday, November 29, 2000 10:17 AM To: DB2-L@RYCI.COM Subject: A simple ZPARM Question [...]
4073 50 27_Re: A simple ZPARM Question12_Moreno, Jose19_Jose_Moreno@BMC.COM31_Wed, 29 Nov 2000 10:46:08 -0600635_iso-8859-1 You're right... But there is a thrird party vendor solution to adjust it "on the fly". Guess from who... http://www.bmc.com/products/proddocview.html?id=2076&r=Americas&l=en
Jose Moreno BMC Software EMEA DB2 Solutions Manager
-----Message d'origine----- De : Richard Arnold [mailto:Richard.Arnold@CHASE.COM] Envoyé : mercredi 29 novembre 2000 17:17 À : DB2-L@RYCI.COM Objet : A simple ZPARM Question
All,
From the manuals and it appears that size of the EDM pool cannot be adjusted "on the fly", and that it requires the ZPARM to changed and re-assembled, and the subsystem bounced. [...]
4124 51 27_Re: A simple ZPARM Question14_Richard Yevich17_ryevich@YAHOO.COM31_Wed, 29 Nov 2000 09:13:11 -0800573_us-ascii Through V6 you are correct. However, one of the more than 60+ new online DSNZPARMS in V7 is the one for altering the size of the EDM POOL.
--- Richard Arnold wrote: > All, > > >From the manuals and it appears that size of the EDM pool cannot > be > adjusted > "on the fly", and that it requires the ZPARM to changed and > re-assembled, > and the > subsystem bounced. > > Can someone verify that I have this right, or correct my if not? > > Thanks in advance. > > ================================================ > To [...]
4176 66 41_Re: parallelism on UNIX - DB2 UDB EE (E?)12_tim malamphy20_timalamphy@YAHOO.COM31_Wed, 29 Nov 2000 09:35:41 -0800574_us-ascii Yes. In loose MVS terms, think of EEE as the parallel-sysplex-data-sharing version, which scales across machines, and EE as the "normal" version.
--- Aurora Dell'Anno wrote: > hi list, > > I am asking this question on behalf of a colleague: > > he has been set the task of finding out, whether > setting up parallelism > with DB2 UDB EE (not sure about the number of E-s > here) is feasible, > impacts on performance if so, how to go about it > etc. > > I am afraid I don't always read matters relating to > [...]
4243 45 26_DB2 vs. FoxPro performance10_Mike Wiles14_mwiles@UCG.COM31_Wed, 29 Nov 2000 12:54:16 -0500562_windows-1252 Hi List!
In my shop, we do a lot of work with FoxPro, Delphi, and DB2 V6. Our DB2 installation is on an RS/6000 H80 running AIX, and seems to be very fast for _almost_ everything we do with it.
However, doing any kind of transformation server side is a problem. What I want to do is get to table Y by querying tables A, B, and C. This involves creating lots of temp tables, inserting the results of queries, querying the created tables, creating more tables, etc. The whole process for a particular report takes about 30 minutes. [...]
4289 105 15_Re: CURRENTDATA13_Mohan, Ashish21_Ashish.Mohan@NIKE.COM31_Wed, 29 Nov 2000 09:59:41 -0800627_iso-8859-1 Lock avoidance can only be used for applications bound with Isolation(CS) and CURRENTDATA(NO) that use read only cursors (cursors 'for fetch only', ORDER BY etc.).
Once that these applications are 'candidates' for lock avoidance, DB2 uses CLSNs (comparing oldest URID over an object with PGLOGRBA over that page) or PUNC (Possibly uncommitted bits) to ensure that it can indeed implement lock avoidance. For most part, I think you can only make an application a 'candidate' for lock avoidance; whether or not DB2 actually goes for it is circumstantial as it depends on the run time usage of that object. [...]
4395 67 23_Select * in Application16_Brookman, Gerald32_gerald.brookman@US.ORIGIN-IT.COM31_Wed, 29 Nov 2000 13:05:38 -0500325_iso-8859-7 I fixed a similar problem in one of our massive applications. They did not want to change all the vendor code from "SELECT * FROM TABLE" because the vendor was extinct and there would NEVER be any changes to the number of columns from the vendor. Then, of course, an end-user decided to customize the system. [...]
4463 50 37_Re: WLM Stored Procedure VARY command16_Pearson, Eric L,23_Eric.Pearson@NSCORP.COM31_Wed, 29 Nov 2000 13:00:51 -0500330_iso-8859-1 Bill, This is an old-fashioned way but does not take any unusual tricks: Have the commands embedded in production batch jobs which you could schedule on demand. The production jobs could have a USER= which had no TSO access but did have ability to do the VARY.
regards, Eric Pearson NS ITO Database Support [...]
4514 51 19_Re: Platinum Course9_Rob Crane22_racrane@CONCENTRIC.NET31_Wed, 29 Nov 2000 11:13:46 -0700345_us-ascii CA has both formal education on RC/Migrator and Detector as well as assistance and training offered via the CA DB2 SE's, formerly the Platinum group that would go on-site and assist with installing the products and getting them set up, etc. Ask the person who did the deal how many days of training were included in the agreement. [...]
4566 24 26_Re: IBM Interpreter Needed15_Smith, Dennis J22_dennis.j.smith@EDS.COM31_Wed, 29 Nov 2000 12:12:48 -0600608_iso-8859-1 -----Original Message----- From: Kornblum, Kenneth [mailto:Kenneth_Kornblum@BMC.COM]
DB2 knows how to deal with this "inconsistency."
DB2 modifies the DBD to reflect the fact that the table space has different size rows within the pages. This flag is the source if the issue. When you create a new table, this flag is not set. DSN1COPY just moves pages into the new table. I forget the error you get (it's been a long time since I seen this situation), but you will know about it when you try to access or create indexes. However, a create index might work if the [...]
4591 98 19_Re: Platinum Course16_Pearson, Eric L,23_Eric.Pearson@NSCORP.COM31_Wed, 29 Nov 2000 13:18:07 -0500478_iso-8859-1 Lisa, Both Detector and Migrator have Tutorials. The one for Migrator is actually pretty good!
regards, Eric Pearson NS ITO Database Support
-----Original Message----- From: Lisa Westcott-Dryer [mailto:lisa_westcott_dryer@FOREMOST.COM] Sent: Wednesday, November 29, 2000 9:55 AM To: DB2-L@RYCI.COM Subject: Platinum Course
A corporate decision has been made to move from BMC to Platinum products. Definately not my decision. [...]
4690 45 11_CURRENTDATA12_Chris Munson18_cmunson@US.IBM.COM31_Wed, 29 Nov 2000 10:36:38 -0800377_us-ascii The CURRENTDATA option only applicable to ISOLATION(CS). ISOLATION(UR) already acquires no locks (except mass delete) so lock avoidance as understood by CURRENTDATA is a moot point. In your example if PROGRAM-A is reading data with ISOLATION(UR), PROGRAM-B is already allowed to update the data selected by PROGRAM-A as a result of the ISOLATION(UR) bind option. [...]
4736 37 9_Question.14_subrata mondal25_subratamondal@HOTMAIL.COM31_Wed, 29 Nov 2000 18:55:54 -0000
4774 40 35_What Does DB2 Reorg pass to DFSORT?13_Lockwood Lyon19_Lockwool@MEIJER.COM31_Wed, 29 Nov 2000 14:29:37 -0500361_US-ASCII As a follow-up to our previous problem, we have the following. (DB2 V6).
We create a single table in a segmented tablespace. One (unique) index on the table. We load data (2,169,480 records) into the table. We run RunStats on the Tablespace. CARD and CARDF both are = 2,169,480. (Note in passing that RunStats were *not* run on the index.) [...]
4815 27 23_DB2 for OS390 Version 613_Ricardo Gomes30_ricardo.gomes@MELLO-IMPERIO.PT31_Wed, 29 Nov 2000 20:18:09 -0000447_- I've heard, unofficially and non-documented, some colleagues of mine talking about V6.2... Still I haven't found such documentation... Is anyone aware on such a new release or does it (they) mean V6 APARs moving it closer to V7 features?
TIA
Best Regards, Cumprimentos,
Ricardo Gomes Servibanca SISTEMAS CENTRAIS Tel.:(351) 21 421 92 74 Fax:(351) 21 421 92 79 Email: ricardo.gomes@mello-imperio.pt ricardo.gomes@bcp.pt [...]
4843 18 4_help10_Elsie Tama24_elsie.m.tama@VERIZON.COM31_Wed, 29 Nov 2000 15:20:03 -0500413_us-ascii What is the down side, or disadvantages, when a table with very long field of primary key( eg. say 23 to 50 bytes long) is been acces.
Thanks,
Elsie
================================================ To change your subscription options or to cancel your subscription visit the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list can be reached at DB2-L-REQUEST@RYCI.COM.
4862 38 39_Re: What Does DB2 Reorg pass to DFSORT?15_Smith, Dennis J22_dennis.j.smith@EDS.COM31_Wed, 29 Nov 2000 14:28:12 -0600560_iso-8859-1 -----Original Message----- From: Lockwood Lyon [mailto:Lockwool@MEIJER.COM]
We run a DB2 Reorg using dynamic sort allocation (SORTNUM parameter), and check the DFSORT messages. Apparently, DFSORT was told to use a file size estimate of FILESZ=E0003001182. (That's 3,001,182) when allocating work space.
Question: Where on earth is it getting this number (the 3,000,000-ish) one? Wouldn't you think that DB2 Reorg would pass the CARD/CARDF as the estimate? Alternatively, is there some formula used to generate this number? [...]
4901 76 39_Re: What Does DB2 Reorg pass to DFSORT?13_Lockwood Lyon19_Lockwool@MEIJER.COM31_Wed, 29 Nov 2000 15:38:11 -0500547_US-ASCII Thanks for pointing out the manual reference. Regrettably (?!) all the table columns are fixed-length. Oh, and Yes, Compress is on. Oddly, in some cases, even though "REORG doubles the estimated FILSZ so that DFSORT allocates enough space ..." we still have cases where the estimate is too small.
Another weird thing is that DFSORT is told (by Reorg, I presume) that the record format is variable. Odd/weird because DB2 should know that the row is fixed-length when uncompressed because of the table column definitions. Sigh. [...]
4978 47 8_Re: help16_Pearson, Eric L,23_Eric.Pearson@NSCORP.COM31_Wed, 29 Nov 2000 16:01:31 -0500477_iso-8859-1 This is mostly a tuning (PCTFREE, REORG frequency...) issue. The main thing with a Primary Key is that if at all possible it be a true Natural Key appropriate for the Entity.
The only downsides specific to DB2: 1) Partitioning can be done only on the first 40 bytes of the clustering index 2) Some catalog statistics are based only on the first 8 bytes of the index 3) Fewer keys per leaf page and the implications such as possibly more index levels.... [...]
5026 55 15_Re: CURRENTDATA10_DANIEL CYR19_DCYR@EMAIL.USPS.GOV31_Wed, 29 Nov 2000 16:08:12 -0600751_iso-8859-1 Murali,
There is an excellent "Red Book" that explains what most of the bind options do. Especially if you are data sharing shop.
SG24-4611-00 Data Sharing Performance Topics
______________________________ Reply Separator _________________________________ Subject: CURRENTDATA Author: owner-db2-l@RYCI.COM at INTERNET Date: 11/28/00 12:41 PM
Can somebody enlighten me with CURRENTDATA and Isolation combination.
Would it be beneficial to use CURRENTDATA(NO) and UR, to take advantage over lock avoidance both ways. For example; with UR, when PROGRAM-A is allowed to read uncommitted data of PROGRAM-B, will CURRENTDATA(NO) allow PROGRAM-B to update the records selected by PROGRAM-A? [...]
5082 96 39_Re: What Does DB2 Reorg pass to DFSORT?12_Bob Lawrence21_rlawrence@BOSCOVS.COM31_Wed, 29 Nov 2000 16:17:22 -0500392_us-ascii Lockwood,
Just a stab in the dark on compression. Compression ratio significantly higher than 50% possibly closer to 85 -90% may cause the doubled estimate to be low. Possibly maintains the fact that the row is compressed and therefore variable length even though it may expand it in the unload
HTH
Bob Lawrence DBA Boscov's Dept Stores Inc [...]
5179 41 30_Re: FW: IBM Interpreter Needed10_Linda Ball18_linda_ball@BMC.COM31_Wed, 29 Nov 2000 15:19:37 -0600553_- I think I can explain it.
When you issue ALTER ADD COLUMN, a value is set in the DBD that accommodates the fact that all your current rows are missing the value for that column. This value makes your rows variable (even if they have only fixed columns). As you INSERT rows, your column value (or the default) will be placed in the new rows. If you SELECT a row without the column, then the default value will be returned. If you define an identical table (including the new column) somewhere and attempt to DSN1COPY the data from the data [...]
5221 83 12_question????29_Sherri Sterling DIGEST NOMIME29_Sherri.Sterling@BPD.TREAS.GOV31_Wed, 29 Nov 2000 16:17:49 -0500727_us-ascii ----- Forwarded by Sherri Sterling/OIT/BPD on 11/29/2000 04:24 PM -----
Sherri Sterling 11/29/2000 04:16 PM
To: Larry Ballard/OIT/BPD@BPD, Scott Cozzens/OIT/BPD@BPD, Dean Garlow/OIT/BPD@BPD, Richard States/OIT/BPD@BPD, Thomas Swanson/OIT/BPD@BPD cc: Subject: question????
Everytime the mainframe db2 os/390 version 6 is bounced or the mainframe is IPLed, the first db2 connect version 6.1 thread established gets a SQL30081N error message. The transaction is reran and a new thread is established and everything is ok. IBM has a problem reported, IY08557 but no fixes yet. We do not want to tell our users to resummit their job everytime they get a SQL30081N message. I would [...]
5305 84 39_Re: What Does DB2 Reorg pass to DFSORT?21_Smith, Ronald M - CNF20_smith.ronald@CNF.COM31_Wed, 29 Nov 2000 13:39:07 -0800613_- I believe you will find that if the tablespace is compressed, DB2 now considers the row data to be variable, since the physical length of the rows will depend on their data and their mapping to the compression dictionary.
> -----Original Message----- > From: Lockwood Lyon [SMTP:Lockwool@MEIJER.COM] > Sent: Wednesday, November 29, 2000 12:38 PM > To: DB2-L@RYCI.COM > Subject: Re: What Does DB2 Reorg pass to DFSORT? > > Thanks for pointing out the manual reference. Regrettably (?!) all the > table columns are fixed-length. Oh, and Yes, Compress is on. Oddly, in > some cases, even though "REORG [...]
5390 64 27_Re: A simple ZPARM Question14_Grainger, Phil20_Phil.Grainger@CA.COM31_Wed, 29 Nov 2000 16:41:08 -0000360_iso-8859-1 Well, until Version 7 anyway. Then you'll be able to refresh Zparms and (hopefully) EDMPOOL size will be one thing you can change.
Of course, if this is an issue today and you can't wait then there are always third-party tools to help (24x7 from CA, Opertune from BMC)
Phil Grainger Product Manager, DB2 Tools Computer Associates [...]
5455 74 17_Re: Gaps in Dates11_Suresh Sane21_data_arch@HOTMAIL.COM31_Wed, 29 Nov 2000 16:08:09 -0600376_- Subrata:
Here is a possible solution.
SELECT A.END_DATE + 1 DAY , B.START_DATE - 1 DAY FROM TRNTRT.JUNK01 A , TRNTRT.JUNK01 B WHERE A.END_DATE + 1 DAY <= B.START_DATE - 1 DAY AND B.START_DATE = (SELECT MIN(START_DATE) FROM TRNTRT.JUNK01 C WHERE C.START_DATE > A.END_DATE)
I am assuming there are no overlaps in dates currently. With the data like: [...]
5530 27 27_DB2 CONNECT ENTERPRISE HELP17_Rabindra Senapati21_rsenapati@HOTMAIL.COM31_Wed, 29 Nov 2000 16:08:50 -0600393_- Hello group,
We have DB2 SUN 7.1 and DB2 VSE 7.1. We want to connect both boxes using DB2 CONNECT ENTERPRISE. We installed DB2 CE in the sun box also. The next step configuration is abstart to us. We do not know what to do in which side( black box and unix box). If any one could help us that will be really great. Please try to help on this problem which is really very urgent. [...]
5558 77 6_Notice14_Philip Gunning24_Philip.Gunning@QUEST.COM31_Wed, 29 Nov 2000 14:38:54 -0800418_- I wanted to let everyone know that I have accepted a position with Quest Software as a Sr. Systems Consultant providing pre-sales support for Quest Central and other future products, with an effective date of 12/4/00. As an Associate List Owner, I will remain fair and impartial on the list. As we all know the list is a "no spin" zone. Feel free to contact me offline if you have any specific questions. -Phil [...]
5636 35 39_Re: What Does DB2 Reorg pass to DFSORT?15_Smith, Dennis J22_dennis.j.smith@EDS.COM31_Wed, 29 Nov 2000 16:42:21 -0600655_iso-8859-1 -----Original Message----- From: Lockwood Lyon [mailto:Lockwool@MEIJER.COM]
Another weird thing is that DFSORT is told (by Reorg, I presume) that the record format is variable. Odd/weird because DB2 should know that the row is fixed-length when uncompressed because of the table column definitions. Sigh.
As most folks have pointed out already, compress rows are variable length rows to DB2. Also, you need to remember that the row length being sent to SORT also contains the largest key length. The size of a row for sorting is the combination of the largest table length plus the largest key length, rehashed [...]
5672 82 30_Re: DB2 vs. FoxPro performance0_22_Rohn.Solecki@MTS.MB.CA31_Wed, 29 Nov 2000 17:11:17 -0600503_us-ascii Well, since no experts have spoken up, I'll take a shot at it, no answers but maybe a question that will get you to try a different approach . . .
- Do you really need the temp table(s)? - Is the delay a result of creating/deleting the temp tables? (If yes, could you create a physical table that is used as a temp table, and cleaned up at the end of each transaction? - Could you rewrite the SQL so that it is either a 3 table join (equalities in WHERE clause) or uses subselects? [...]
5755 122 41_Re: parallelism on UNIX - DB2 UDB EE (E?)16_Aurora Dell'Anno26_aurora.dellanno@SISTINF.IT31_Thu, 30 Nov 2000 00:11:39 +0100455_us-ascii Tim,
thanks a bunch for your reply - unbelievable, I finally understood why I was getting so mixed up with all those EEEEEEEs!
I'm copying my colleague in on this, do you think I or him could pester you offline regarding this matter? (the parallelism, not the E's, of course ;-)
ciao!
Aurora Emanuela Dell'Anno Systems Programmer Sistemi Informativi S.p.A. e-mail: aurora.dellanno@sistinf.it tel. +39 335 7429486 [...]
5878 44 19_Re: Platinum Course15_Bell, Raymond W31_Raymond.W.Bell@TEAM.TELSTRA.COM31_Thu, 30 Nov 2000 12:00:01 +1100360_- As far as RC/Migrator is concerned, you'll find it behaves very similarly to Change Manager. I quite like both tools, but find myself tending towards RC/Migrator. That'll probably swing back the other way, now that I'm at a site with the BMC tools again. Anyway, RC/Migrator at least is a good tool. Can't vouch for Detector; haven't used it that much. [...]
5923 68 37_Re: WLM Stored Procedure VARY command14_James Campbell29_James.Campbell@HANCORP.COM.AU31_Thu, 30 Nov 2000 13:17:05 +1100370_iso-8859-1 Do your sysprogs object to you being able to refresh/resume/quiesce the wlm address spaces? (As distinct from all the other groovy things a VARY WLM can do.) If not, their automation tool should be sophisticated enough so that you can enter a TSO command, which the tool intercepts, validates the issuer and then issues the appropriate VARY WLM command. [...]
5992 13 22_testing, please ignore16_Michael Horowitz27_michael@MICHAELHOROWITZ.ORG31_Thu, 30 Nov 2000 01:29:50 -0500279_us-ascii This is just a test posting.
================================================ To change your subscription options or to cancel your subscription visit the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list can be reached at DB2-L-REQUEST@RYCI.COM.
6006 24 45_Trace R Jones/SS/HouInd is out of the office.11_Trace Jones31_Trace_R_Jones@RELIANTENERGY.COM31_Thu, 30 Nov 2000 01:01:23 -0600549_us-ascii I will be out of the office from 11/30/2000 until 12/04/2000.
I will respond to your message when I return.
If you need DB2 assistance, please call:
Robert Lagoudis x77874
If you need Oracle Assistance, please call:
Joey Foushee x5632 Georgia Carson x5870
================================================ To change your subscription options or to cancel your subscription visit the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list can be reached at DB2-L-REQUEST@RYCI.COM.
6031 57 24_IMS and Stored Procedure12_Jacob Ganzel19_jacobg@SEMECH.CO.IL31_Thu, 30 Nov 2000 09:04:32 +0200226_x-user-defined Hello list
Can Store procedure UPDATE IMS database ? Is there any limitations? Please refer to documentations.
Thanks. Jacob. ---------------------------------------------------------------
6089 26 20_JRE for DB2 on Linux13_Viswanathan N33_Viswanathan.N@POWAIMAIL.LTITL.COM31_Thu, 30 Nov 2000 13:32:31 +0530544_us-ascii hi ! ! I successfully installed DB2 V7.1 on Linux ..intel platform ... just waiting for the S/390 Platform to reach our site. Well I am not able to configure JRE for running control center on the Linux. the DB2CC shell script provided by IBM in /usr/IBMdb2/V7.1/bin directory has the JRE options as " jre JRE_OPTIONS " . now . I downloaded JRE for Linux . but that package does not have jre executable . Its creating the JRE directory and all the other stuff . but the jre executable is not there .. . Which package will give the [...]
6116 175 39_Re: What Does DB2 Reorg pass to DFSORT?0_18_mebert@AMADEUS.NET31_Thu, 30 Nov 2000 09:26:24 +0100390_us-ascii There is a PTF available for DB2 V5 at least (maybe V6 also; I forgot the number) that causes DB2 to take PAGESAVE into account. If your tablespaces have PAGESAVEs>50%, this is quite vital if you use dynamic allocation. I didn't understand the estimates passed to DFSORT either (before or after the PTF application), it's probably one of those things you just have to accept. [...]
6292 57 28_Re: IMS and Stored Procedure14_James Campbell29_James.Campbell@HANCORP.COM.AU31_Thu, 30 Nov 2000 20:03:09 +1100492_x-user-defined It appears that the answer is yes - "IMS commits or backs out the work as part of RRS two-phase commit when the stored procedure caller executes COMMIT or ROLLBACK." (Application Programming and SQL Guide both V5 and V6).
Never having done it, I don't know about limitations.
It took me all of, umm, about 30 seconds to find this quote. /* standard disclaimer */ James Campbell DBA Hansen Corporation, Doncaster +61 3 9843 8442 James.Campbell@Hancorp.com.au [...]
6350 49 15_Re: DDF Failure11_David Ayers24_david.ayers@HIGHMARK.COM31_Wed, 29 Nov 2000 11:42:01 -0500568_us-ascii dan, look to see if there are any messages in your db2mstr / db2ddf addr spaces
Daniel Sullivan @RYCI.COM> on 11/29/2000 11:19:24 AM
Please respond to DB2 Data Base Discussion List
Sent by: DB2 Data Base Discussion List
To: DB2-L@RYCI.COM cc:
Subject: DDF Failure
A co-worker is trying to go from one DB2 on one lpar to another DB2 on another lpar using Websphere JAVA and DDF. She is receiving the following error: [...]
6400 21 30_Re: Runstats problems DB2 V5.116_Martin Berchtold31_Martin.Berchtold@SYSTEMATICS.DE31_Thu, 30 Nov 2000 04:26:02 -0600450_- Hi Rick,
I tried in the meantime to get an access to the recommended IBM INFOLINK, but it's a very triggy action for us in Europe. We don't have the same possibilities as you have in the states or somewhere else, in my opinion it's a joke. Now my question - you mentioned that you got 10 - 15 hits for DSNUSIDX+0316. Is it possible to send me this list - or at least the APAR numbers? Thanks in advance - even if it is is not possible. [...]
6422 61 30_Please, help on Signoff !!!!!!6_monica23_655.monica@BANEB.COM.BR31_Thu, 30 Nov 2000 09:32:45 -0200347_iso-8859-1 Hi, list,
I really need some help on how to signoff. Our company decided to change all e-mail accounts, and, since the new one became available, the last one was deleted, and cannot be used anymore. I need to signoff my old e-mail account, but I can't use this account to send a SIGNOFF DB2-L. What should I do to signoff ? [...]
6484 28 34_Re: Please, help on Signoff !!!!!!13_Seibert, Dave26_Dave_Seibert@COMPUWARE.COM31_Thu, 30 Nov 2000 06:56:01 -0500632_iso-8859-1 Hi Monica, You might try signing off from the web page:
http://jupiter.ryci.com/cgi/wa.exe?SUBED1=db2-l&A=1
You might also send your original message to the list owners at
DB2-L-request@ryci.com
One of those should resolve your problem.
Good luck.
David Seibert Compuware Corporation File-AID Product Architect Dave.Seibert@Compuware.com
================================================ To change your subscription options or to cancel your subscription visit the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list can be reached at DB2-L-REQUEST@RYCI.COM.
6513 25 21_surprising result set15_Walter Janissen27_walter.janissen@VICTORIA.DE31_Thu, 30 Nov 2000 06:15:30 -0600558_- Hi List
I don't know, if anyone does expect an empty result set running the follwing SQL:
SELECT VNR, RR_RV_BEST_KENN FROM D_ABL_1 A WHERE VNR NOT IN (SELECT VNR FROM D_ABL_2) ORDER BY RR_RV_BEST_KENN, VNR
Well following information: Both tables contain some hundred rows, and no VNR of D_ABL_1 is contained in D_ABL_2, so this SQL is very straight forward, but no row will be returned. How can that be? The answer is, that table D_ABL_2 has one row with a NULL value in VNR. I'm still wondering, if the result is really correct? [...]
6539 57 31_trigger package and access path15_Walter Janissen27_walter.janissen@VICTORIA.DE31_Thu, 30 Nov 2000 06:22:29 -0600668_- Hi
I just created the following insert trigger:
CREATE TRIGGER DB2.VATI0025 AFTER INSERT ON DB2.VATB0025 REFERENCING NEW AS N FOR EACH ROW MODE DB2SQL WHEN(N.BEARB_GRD <> '05') BEGIN ATOMIC INSERT INTO DB2.VATB0026 ( ZUG_ID , INSERT_TIMESTAMP , ZUG_STS , HERK_KZ , START_KW , H_KW ) VALUES ( N.ZUG_ID , CURRENT TIMESTAMP , N.ZUG_STS_SL , 'H' , CASE N.KW_GENERIER_KZ WHEN 'X' THEN SUBSTR(DIGITS( MICROSECOND(N.BEARB_ZP) ), 5, 6) ELSE ' ' END , CASE N.KW_GENERIER_KZ WHEN 'X' THEN SUBSTR(DIGITS(1875423 * DECIMAL(N.HOST_KW)), 11, 8)!! SUBSTR(DIGITS(8745261 * DECIMAL(N.HOST_KW)), 11, 8) ELSE ' ' END ) ; END# Then I rebound the trigger package with [...]
6597 19 33_Unload with DSNTIAUL under DB2 V50_25_heiko.schmaelzle@SI-BW.DE31_Thu, 30 Nov 2000 15:02:04 +0100471_us-ascii Hi List!
Which rows will be unloaded by this utility if there are still changed rows of the table in the bufferpool (the application program already said commit). Are the changes included?
TIA H. Schmaelzle
================================================ To change your subscription options or to cancel your subscription visit the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list can be reached at DB2-L-REQUEST@RYCI.COM.
6617 34 34_What happens when you COLD start ?12_Kirk Hampton16_khampto1@TXU.COM31_Thu, 30 Nov 2000 08:14:10 -0600540_us-ascii Oh wise ones: We are running DB2 UDB v6.1 on OS/390 2.8 at the PUT0009 maint level. We had a dynamic UPDATE SQL encounter a bug which crashed our test sybsystem. This left one PART of a tablespace in RESTP (Restart Pending), also one PART of the partitioning index and an NPI are all in RESTP. I had a similar problem back when we were on the May refresh level, and IBM's recommendation then was to COLD start my DB2 to clear the RESTP flags. Instead, since it was a test tablespace, we DROPed it and rebuilt it from backups. [...]
6652 100 39_Re: What Does DB2 Reorg pass to DFSORT?13_Lockwood Lyon19_Lockwool@MEIJER.COM31_Thu, 30 Nov 2000 09:15:03 -0500542_ISO-8859-7 Thanks to all who wrestled with this one. A few things to keep in mind.
According to the RTFM, DFSORT uses its "E15" exit to invoke DB2 to retrieve records to sort. This makes sense, since only DB2 can read tablespace pages and decompress, and decode the row/column information. By the time DFSORT gets the data it's already been uncompressed and expanded (for example, I *assume* that TimeStamp columns are expanded to 26 bytes, Date columns to 10 bytes, etc.; otherwise, they wouldn't sort correctly ... would they?) [...]
6753 65 57_Re: DB2-L Digest - 26 Nov 2000 to 28 Nov 2000 (#2000-346)14_Ciran McCarthy28_cmccarthy.cs@CLEARSTREAM.COM31_Thu, 30 Nov 2000 15:06:40 +0100683_- Since you sent your 'probe' I have had no further correspondence from you. Could you please investigate. Thanks.
> ---------- > From: Automatic digest processor[SMTP:LISTSERV@RYCI.COM] > Reply To: DB2 Data Base Discussion List > Sent: Tuesday, November 28, 2000 08:00 > To: Recipients of DB2-L digests > Subject: DB2-L Digest - 26 Nov 2000 to 28 Nov 2000 (#2000-346) > > <>< DB2 for OS390 (11/27/00)>><>< Regarding DB2 dataset Allocation>><>< Re: Why is Fisher-Price knowledge re...>><
DSNTIAUL uses dynamic SQL to "unload" the table, so yes any committed data in buffer pools is included!
HTH, Rick Creech
>From: heiko.schmaelzle@SI-BW.DE >Reply-To: DB2 Data Base Discussion List >To: DB2-L@RYCI.COM >Subject: Unload with DSNTIAUL under DB2 V5 >Date: Thu, 30 Nov 2000 15:02:04 +0100 > >Hi List! > >Which rows will be unloaded by this utility if there are still changed rows >of >the table in the bufferpool (the application program already said commit). >Are >the changes included? > >TIA >H. Schmaelzle > >================================================ >To change your subscription options or to cancel your subscription visit >the DB2-L [...]
6864 10 37_Re: Unload with DSNTIAUL under DB2 V515_Walter Janissen27_walter.janissen@VICTORIA.DE31_Thu, 30 Nov 2000 08:39:38 -0600247_- yes
================================================ To change your subscription options or to cancel your subscription visit the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list can be reached at DB2-L-REQUEST@RYCI.COM.
6875 53 38_Re: What happens when you COLD start ?14_Grainger, Phil20_Phil.Grainger@CA.COM31_Thu, 30 Nov 2000 14:44:35 -0000423_iso-8859-1 I only ever cold-started once, to prevent DB2 trying to back out units of work that were on archive logs we didn't have any more - (oops).
Cold starting basically says to DB2 "before this RBA nothing ever happened that you need to know about". As a result, one thing you should do after a cold start is image copy EVERYTHING IN SIGHT as you cannot recover back to the 'other side' of the cold start. [...]
6929 37 39_Re: What Does DB2 Reorg pass to DFSORT?15_Smith, Dennis J22_dennis.j.smith@EDS.COM31_Thu, 30 Nov 2000 09:00:33 -0600468_ISO-8859-7 -----Original Message----- From: Lockwood Lyon [mailto:Lockwool@MEIJER.COM]
For the particular table in question there are *no* variable-length columns, only Date, Timestamp, Integer, Char, and Decimal. Some (seven) columns are nullable. So, one might *assume* that Reorg "knows" this (it's got the table RecLength and column definitions in the catalog, after all) and would pass a "Record Format is Fixed" to DFSORT. I guess not. [...]
6967 19 38_Re: What happens when you COLD start ?10_Max Scarpa16_mscarpa@CESVE.IT31_Thu, 30 Nov 2000 15:58:46 +0100508_us-ascii I heard only in this last year the words 'Cold Restart' for DB2 (I must admit my ignorance, here as in many & many others topics...)
Is it a equivalent to the old CRC restart ? They sounds very similar and to avoid as plague.
Thanks Max Scarpa
================================================ To change your subscription options or to cancel your subscription visit the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list can be reached at DB2-L-REQUEST@RYCI.COM.
6987 83 39_Re: What Does DB2 Reorg pass to DFSORT?13_Lockwood Lyon19_Lockwool@MEIJER.COM31_Thu, 30 Nov 2000 10:29:04 -0500509_ISO-8859-7 Thanks, Dennis. Yes, we DBAs are finicky. *But* we are also the customer.
As for fixed versus variable and nulls ¯ I hear you. I investigated. Only one table in the segmented tablespace, only one index. As for the nulls, ... hmmm. How would DB2 be able to figure out what was data and what was null if one-byte null indicators weren't there for *every* field? It doesn't matter, of course. How Reorg works (?!) isn't my concern, and maybe the "V" that DFSort gets isn't that important. [...]
7071 229 39_Re: Does an elephant fly? No? Prove it.14_Laredo, Andrew19_Andy_Laredo@BMC.COM31_Thu, 30 Nov 2000 09:37:16 -0600413_iso-8859-7 I think you all are not addressing his problem/question....
- the SELECT * are coded - in several/many application programs
Management wants proof that they have to go back and re-code/fix the programs. The fact that SEL * is not a good idea, or generates a warning instead of error, etc. is not proof to management that it needs re-coding (in my opinion, and probably theirs also). [...]
7301 98 24_Re: JRE for DB2 on Linux15_Blaney, Bill L.21_Bill.L.Blaney@SSA.GOV31_Thu, 30 Nov 2000 10:43:18 -0500445_iso-8859-1 Vishy,
I've just started playing with DB2 UDB v6.1 for Linux, and ran into some similar problems. Here's what I did to get db2cc to work:
First, get a copy of the article Paul Zikopoulos wrote in the May, 2000 edition of Linux Journal. You can a find a copy of this on the web at http://www2.linuxjournal.com/lj-issues/issue73/3937.html It's written for UDB v6.1, but I believe that most of it is still applicable. [...]
7400 49 37_Re: Unload with DSNTIAUL under DB2 V515_Toine Michielse18_vndobtm@US.IBM.COM31_Thu, 30 Nov 2000 16:42:16 +0000507_us-ascii Hello Heiko,
Yes, DSNTIAUL (a program not a utility) just uses normal SQL so changes will included even if the pages have not been externalized to DASD.
Regards, Toine Michielse
heiko.schmaelzle@SI-BW.DE@RYCI.COM> on 11/30/2000 02:02:04 PM
Please respond to DB2 Data Base Discussion List
Sent by: DB2 Data Base Discussion List
To: DB2-L@RYCI.COM cc: Subject: Unload with DSNTIAUL under DB2 V5 [...]
7450 71 10_Re: Notice10_Mark Labby16_mlabby@PHEAA.ORG31_Thu, 30 Nov 2000 11:02:49 -0500443_iso-8859-1 I'd just like to again offer Phil my congratulations on his new position and thank him for the time that he spent with us at PHEAA.
Quest Software has made a very good hire and I'm sure you will do very well for them.
Good luck, Phil!!
Philip Gunning cc: Sent by: DB2 Subject: Notice Data Base Discussion List [...]
7522 24 37_diff. b/w catridge tape and 3480 tape17_ravi kumar hassan18_ravibh@HOTMAIL.COM31_Thu, 30 Nov 2000 10:09:13 -0600271_- Hi Folks,
In our shop we are using a catridge and 3480 tape for taking a backups but now they are asking us to stop taking backups using catridge tape and use only 3480 tapes but i am not able to understand what is the diff. b/w catridge and 3480 tape..... [...]
7547 41 42_Group bufferpool and CA/Platinum DB2 tools12_Richard Pack23_rpack1@MAIL.STATE.MO.US31_Thu, 30 Nov 2000 10:09:30 -0600365_- We have set up our first data sharing group with two members. I would like to know how others, that have the CA/Plainum DB2 tools, have chosen to allow them to be used.
It seems that if I want to use the tools from either member, I must have a group bufferpool defined the the bufferpools the TS's and IX's are in. This seems like a waste of storage. [...]
7589 42 53_Re: Why is Fisher-Price knowledge required for OS/39012_Jim Harrison17_jharrison@GMX.NET31_Thu, 30 Nov 2000 11:05:33 -0500546_us-ascii I'm not IBM and I can't tell you why, but you might be interested in this offer while it lasts:
>Free Download of DB2 Computer Based Training now through December 15, > 2000 > >Are you a DBA who is already proficient in one or more nonIBM >relational database products, but could use some help in coming >up to speed with DB2 UDB? IBM Learning Services has a course >designed to give you a fast path to build the needed skills in as >little as eight hours. As part of a special offer in conjunction >with the trial copies of [...]
7632 41 41_Re: diff. b/w catridge tape and 3480 tape12_MARTIN WOLFF31_martin_wolff@GLOBALCROSSING.COM31_Thu, 30 Nov 2000 11:26:30 -0500368_us-ascii Hi Ravi,
The term cartridge is now generic. It can mean 3480 or 3490. It may be that your people are asking you to only use the smaller 3480s (UNIT=3480) and not the bigger 3490s which may be your installation default for 'UNIT=CART' or similar.
Or alternatively, just ask them!
Martin Wolff.
ravi kumar hassan wrote: [...]
7674 209 38_Re: What happens when you COLD start ?13_John Arbogast32_john.arbogast.cquz@STATEFARM.COM31_Thu, 30 Nov 2000 10:19:00 -0600501_iso-8859-1 Can't you recover "tocopy" on the 'other side'? We have had to cold start to get objects out of AREST status. Nice feature of V6?!
-----Original Message----- From: Phil.Grainger@CA.COM [mailto:Phil.Grainger@CA.COM] Sent: Thursday, November 30, 2000 8:45 AM To: DB2-L@RYCI.COM Subject: Re: What happens when you COLD start ?
I only ever cold-started once, to prevent DB2 trying to back out units of work that were on archive logs we didn't have any more - (oops). [...]
7884 74 41_Re: diff. b/w catridge tape and 3480 tape12_Kirk Hampton16_khampto1@TXU.COM31_Thu, 30 Nov 2000 11:30:01 -0600574_us-ascii I believe Martin is correct, all 3480, 3490, and 3590 tapes are a cartridge format, but your systems folks are probably referring to the use of the word CART in the UNIT=CART parameter in your JCL. CART is a generic device type and is installation defined (can be different at every shop). They are probably just wanting you to make all your backup JCL use the same specification. CART at your shop may mean the job can use whatever scratch tape is available, so you could be writing a small dataset to a high-capacity 3590 cartridge and wasting its capacity. [...]
7959 49 53_Re: Why is Fisher-Price knowledge required for OS/39013_Jimmy Willett24_Jimmy.Willett@NCMAIL.NET31_Thu, 30 Nov 2000 12:56:04 -0500524_us-ascii tried to get on the page, but got page missing error...
Jim Harrison wrote:
> I'm not IBM and I can't tell you why, but you might be interested in this > offer while it lasts: > > >Free Download of DB2 Computer Based Training now through December 15, > > 2000 > > > >Are you a DBA who is already proficient in one or more nonIBM > >relational database products, but could use some help in coming > >up to speed with DB2 UDB? IBM Learning Services has a course > >designed to give you a fast path [...]
8009 21 18_? db2 unix version11_David Ayers24_david.ayers@HIGHMARK.COM31_Thu, 30 Nov 2000 13:13:48 -0500457_us-ascii is there a way to get the current version + ptf that a specific db2 udb instance running on sun is using. We are running multiple fixpacks (v 6.1) on the same servers. My understanding of the DB2LEVEL command is that it will only display what the base code + latest ptf applied. When we run this command all instances show the same information. Am I mistaken or is there another way to display what version+ptf is applied for a given instance [...]
8031 72 46_Re: Group bufferpool and CA/Platinum DB2 tools11_rick creech18_ykcirc@HOTMAIL.COM31_Thu, 30 Nov 2000 12:14:33 -0600563_- Hello,
I set up a data sharing system at a site with Platinum tools and I made sure that the tools were accessible from both members. Also, I decided it to be a good default to have group buffer pools for all virtual buffer pools, thinking that would give maximum flexibilty for anything to be able to run anywhere. This caused no problems...and it kept me from having the problem you encountered. Of course, you must have the coupling facility capacity and you must size your structures reasonably. We had 1 gig in each of two coupling facilities. [...]
8104 30 53_Re: Why is Fisher-Price knowledge required for OS/39012_Jim Harrison17_jharrison@GMX.NET31_Thu, 30 Nov 2000 13:37:00 -0500548_us-ascii Sorry - somehow a some dashes got dropped when I cut & pasted the note. It should be:
http://isource.ibm.com/cgi-bin/goto?on=db2-cbt
which then promptly redirects you to this page (in case the above still doesn't work)
http://www6.software.ibm.com/dl/dbas/db2udbexp-p
At 12:56 PM 11/30/00 -0500, Jimmy Willett said: >tried to get on the page, but got page missing error... > >Jim Harrison wrote: > > > I'm not IBM and I can't tell you why, but you might be interested in this > > offer while it lasts: > > [...]
8135 19 34_clustered versus Unclustered index10_Elsie Tama24_elsie.m.tama@VERIZON.COM31_Thu, 30 Nov 2000 14:54:38 -0500441_us-ascii What happens if an index is not Clustered? Please let me know the advantages and disadvantages of this. Any suggestioins will be greatly appreciated.
Thanks,
Elsie 703-974-2605
================================================ To change your subscription options or to cancel your subscription visit the DB2-L webpage at http://www.ryci.com/db2-l. The owners of the list can be reached at DB2-L-REQUEST@RYCI.COM.
8155 88 28_Re: IMS and Stored Procedure12_Myron Miller22_myronwmiller@YAHOO.COM31_Thu, 30 Nov 2000 12:11:56 -0800574_us-ascii You're basically using the DRA interface - somewhat similar to the DRA interface that CICS uses to connect to IMS/DBCTL.
Catch is this requires IMS V6 at the minimum. --- James Campbell wrote: > It appears that the answer is yes - "IMS commits or > backs out the work as > part of RRS two-phase commit when the stored > procedure caller executes > COMMIT or ROLLBACK." (Application Programming and > SQL Guide both V5 and V6). > > Never having done it, I don't know about > limitations. > > It took me all of, umm, about [...]
8244 55 38_Re: clustered versus Unclustered index0_26_truman.g.brown@VERIZON.COM31_Thu, 30 Nov 2000 15:30:35 -0500512_us-ascii Elsie,
Tables are normally "clustered" (physical row sequenced by the clustering "key") to improve performance for sequential processes that use sequential or dynamic prefetch. Non-clustering indexes are usually those used for random access and/or list pre-fetch.
George
"Elsie Tama" on 11/30/2000 02:54:38 PM
Please respond to "DB2 Data Base Discussion List"
To: DB2-L@RYCI.COM [...]
8300 42 38_Re: clustered versus Unclustered index11_Sidney Chan17_sichan@SIEBEL.COM31_Thu, 30 Nov 2000 12:22:36 -0800470_iso-8859-1 If an index is clustering but not clustered, that means the table data is not in the sequence of the clustering index by a certain percentage. I don't recall that percentage off hand, but it is somewhere around 90. Assuming it is beneficial to have the table data in a certain sequence (that's why you choose a clustering index to begin with), the benefit is no longer there when the index is no longer clustered, i.e. the table data is not in sequence. [...]
8343 60 38_Re: clustered versus Unclustered index11_Joseph Link26_joe.link@FIRSTDATACORP.COM31_Thu, 30 Nov 2000 14:44:49 -0600289_us-ascii DB2 allows one index to be defined as "Clustering". This is the index that DB2 will use to physically organize the rows in the table. Other indexes on the table may be considered "Clustered" if a high percentage of the rows in the table are in order by this index ( >90% ?). [...]
8404 180 76_Re: question regarding your use of OS/390 DB2 and the web - secur ity issues12_Myron Miller22_myronwmiller@YAHOO.COM31_Thu, 30 Nov 2000 12:46:08 -0800332_us-ascii Question Gerry: If you use individual IDs, what happens when you get 500,000 users to your system (or assuming big system 50-100 million). Is your system programmer really going to enjoy having a RACF database this size? I strongly suspect not. (Can you imagine the incore RACF table of users with 100 million users.) [...]
8585 54 38_Re: clustered versus Unclustered index15_Billings, Linda30_linda.billings@DOA.STATE.WI.US31_Thu, 30 Nov 2000 14:52:02 -0600388_iso-8859-1 Hi, Elsie,
That depends on the meaning of your question. If you are asking what happens if an index defined as a clustering index is pointing to data that is no longer in cluster order (less than 80%) then list prefetch kicks in. DB2 accesses all the keys that it needs from the index, sorts them into cluster order and then fetches them using sequential prefetch. [...]
8640 134 41_Re: parallelism on UNIX - DB2 UDB EE (E?)11_Jeremy Dodd21_jeremy.dodd@WHICH.NET31_Thu, 30 Nov 2000 21:35:37 +0000358_us-ascii Aurora,
I would put the answer differently in AIX terms - being only AIX based.
EE does allow for parallelism. You need to set the intra parallel db cfg parameter to yes, set max query degree to more than one and make sure the current degree or bind degree is more than one. This then allows queries to be udnertaken in parallel. [...]
8775 49 11_ERQUAL 53CB0_22_aime_Duclos@BCBSME.COM31_Thu, 30 Nov 2000 16:53:59 -0500494_us-ascii I have received the following messages when attempting a Select (*) on each of two tables recently copied using image copies from one subsystem via DSN1COPY to put data in another subsystem. I'm fairly familiar with DSN1COPY an its quirks, as we use it quite a bit. Although I may be missing something (and I've had some other sets of eyes check it out, too) the DSN1COPY seems ok. Both tablespaces are partitioned and have one table each. The image copies used are by partition. [...]
8825 171 41_Re: parallelism on UNIX - DB2 UDB EE (E?)11_Sidney Chan17_sichan@SIEBEL.COM31_Thu, 30 Nov 2000 13:54:57 -0800450_iso-8859-1 I'd just like to point out there is a substantial architecture difference between AIX based and MVS based multi-node parallelism. In AIX, it is a share-nothing architecture where as in MVS it is a share-disks architecture. DB2 sysplex in MVS is much like the Oracle Parallel Server (OPS) in which one database can be mounted to multiple Oracle instances, and in DB2 EEE AIX a given database partition can only be mounted to one node. [...]
8997 31 33_List PreFetch/Sequential prefetch10_Elsie Tama24_elsie.m.tama@VERIZON.COM31_Thu, 30 Nov 2000 17:49:01 -0500294_us-ascii If an index defined as a clustering index is pointing to data that is no longer in cluster order List prefetch, as I understand will set in. IE, DB2 accesses all the keys that it needs from the index, sorts them into cluster order and then fetches them using sequential prefetch. [...]
9029 23 37_Re: List PreFetch/Sequential prefetch0_26_truman.g.brown@VERIZON.COM31_Thu, 30 Nov 2000 17:59:46 -0500384_us-ascii Elsie,
ANY form of prefetch is usually a good thing as pages are read asynchronously before they are needed by the application, eliminating most of the I/O wait time.
Random I/O is synchronous and will incur I/O wait.
Of course I don't know which version of DB2 you have, type of DASD, or amount of DASD cache you have, but that's the general idea. [...]
9053 125 30_Re: DB2 vs. FoxPro performance10_Mike Wiles14_mwiles@UCG.COM31_Thu, 30 Nov 2000 18:17:41 -0500371_windows-1252 Thanks for your comments! I thought that the subject line would incite more of a response, but I guess not!
First, I shouldn't have said that I'm querying tables a, b, and c. There are actually about ten tables involved in this whole process. The final result is a denormalization of these based filtering out records based on date and location. [...]
9179 78 15_Re: ERQUAL 53CB15_Toine Michielse18_vndobtm@US.IBM.COM30_Fri, 1 Dec 2000 00:18:14 +0000383_us-ascii Hello Aime,
I think the erqual indicates a missing column in the data. So for instance, the table in the 'source' DB2 has had a column added. No REORG was done between the alter and the mage copy that you use for DSN1COPY.
I would look along those lines....
Toine Michielse
aime_Duclos@BCBSME.COM@RYCI.COM> on 11/30/2000 09:53:59 PM [...]
9258 78 15_Re: ERQUAL 53CB15_Toine Michielse18_vndobtm@US.IBM.COM30_Fri, 1 Dec 2000 00:22:36 +0000760_us-ascii Aime,
PS. Another reason maybe PQ41076. Release 610 : UQ47020 available 00/10/03 (F010 )
Regards,
Toine Michielse
aime_Duclos@BCBSME.COM@RYCI.COM> on 11/30/2000 09:53:59 PM
Please respond to DB2 Data Base Discussion List
Sent by: DB2 Data Base Discussion List
To: DB2-L@RYCI.COM cc: Subject: ERQUAL 53CB
I have received the following messages when attempting a Select (*) on each of two tables recently copied using image copies from one subsystem via DSN1COPY to put data in another subsystem. I'm fairly familiar with DSN1COPY an its quirks, as we use it quite a bit. Although I may be missing something (and I've [...]
9337 43 30_datalinks installation problem12_tim malamphy20_timalamphy@YAHOO.COM31_Thu, 30 Nov 2000 15:50:52 -0800351_us-ascii One of my coworkers is having a problem with a DATALINKS installation. It's a 7.1 version. I couldn't even begin to find any of the codes documented. I'm thinking a reboot might help the situation, but we're a bit limited in our ability to boot everyone off the system. I didn't see anything on IBM's site. Can anyone point me anywhere? [...]
9381 44 75_Re: question regarding your use of OS/390 DB2 and the web - security issues9_Eric Kwai18_Eric_Kwai@DTCC.COM31_Thu, 30 Nov 2000 20:27:23 -0600581_- Larry: Our shop use Stored Procedure exclusively on the OS/390 for remote client access. Using Stored Procedure on the host can save network traffic and centralize your business logic to achive performance and maintainance objective.
In term of security, you can
1. Use a challange method (have the user enter their userid/password at Application login, depends on your shop's RACF password expiration policy.) Userid can be translated to a secondary authid known only to DDF in the usernames table for ease of grant permission. (Supported by SNA access only) [...]
9426 73 25_Re: surprising result set14_James Campbell29_James.Campbell@HANCORP.COM.AU30_Fri, 1 Dec 2000 14:16:05 +1100548_iso-8859-1 Like you, I find this counter intuitive. However, a re-read of the SQL reference shows that DB2 is behaving correctly.
Basically: - a IN predicate on a null value has an "unknown" truth status. (The IN predicate will be true if another value gives it a "true" status.) - NOT "unknown" is also "unknown" - so NOT IN (... "unknown" ...) is either false (for those rows which match another value in the list) or "unknown" (for every other row) - if the search condition for a row is "unknown" then the row will not be selected. [...]
9500 57 39_Re: What Does DB2 Reorg pass to DFSORT?14_James Campbell29_James.Campbell@HANCORP.COM.AU30_Fri, 1 Dec 2000 14:57:30 +1100479_iso-8859-7 1) DB2 stores data such that their bytewise order matches their sorting order. So timestamps would be left at 10 bytes, dates at 4, etc. (See Admin Guide Appendix "Writing Exit routines", section "Row Formats".)
2) Are you using/considered KEEPDICTIONARY? Advantages include "[e]xisting compressed rows do not have to be compressed again." Implies that DB2 can expand the row, extract the sort key, and then pass the sort key and compressed row to DFSORT. [...]
9558 48 22_Re: ? db2 unix version11_Bruce Allen17_ballen@SPOTON.COM30_Fri, 1 Dec 2000 16:04:23 +1100413_- Hi Dave, Usually there is one copy of a version for DB2 on unix. For DB2/Sun, v6, the software lives in /opt/IBMdb2/V6.1. V7 will probably live in /opt/IBMdb2/V7.1. The sqllib under the instance home directory contains some data, and links to the software libraries. So, the reason db2level tells you it is all one level is that it is probably only expecting a single level (I am not sure where he looks). [...]