1 WWW.IDUGDB2-L.ORG /home/listserv/home/db2-l September 2009, week 4 2 27 35_Which member from Datasharing group11_Sameer Rana21_sameerdrana@GMAIL.COM31_Tue, 22 Sep 2009 16:14:47 +0000303_utf-8 If I submit a batch job on an LPAR A which has two members ( DB3Z and DB5Z ) of a datasharing group DSNZ and I refer to DSN(DB2G) as the generic group attach name in my job and SYSAFF LPAR A.What governs whether the job needs to run on DB3Z or DB5Z ? How can I check which member it ran on ? [...] 30 140 28_Re: identity column db2 unix14_Renzo razzetti24_renzo.razzetti@GMAIL.COM31_Tue, 22 Sep 2009 13:33:48 -0300613_ISO-8859-1 Jack

Thanks for your answer It isn´t exactly what we need.

After do a "db2 alter table xxxdba.t1 alter column eirarintid restart with v1", I want to check the restart value. Where DB2 save the restart value ???

Thanks in advance.

RR



On Mon, Sep 21, 2009 at 1:46 PM, Jack Campbell wrote:

> Renzo > > an IDENTITY column is just another type of sequence object, that is > attached > to a particular table/column > > Here is some sample SQL to retrieve details. In my example I pull all > details > for tables prefixed "DW" > [...] 171 109 58_Re: DB2-UDB fro LUW - restarted identity column next value14_Renzo razzetti24_renzo.razzetti@GMAIL.COM31_Tue, 22 Sep 2009 13:38:05 -0300634_ISO-8859-1 Peter

Have you find how ??? I have the same problem...

Thank you RR

On Fri, Aug 14, 2009 at 4:10 AM, Peter Richardson < peter.richardson@hbf.com.au> wrote:

> Hiya's, > > I'm tryna find the next value for an identity column in the situation where > the > column has been restarted with a particular value. In this case, > SYSIBM.SYSSEQUENCES has the LASTASSIGNEDVALUE as null and > SYSCAT.COLIDENTATTRIBUTES has NEXTCACHEFIRSTVALUE SET to > the original start value (not the restart value). > > I know that under z\os SYSIBM.SYSSEQUENCES has a RESTARTWITH > column, but this does not appear [...] 281 41 39_Re: Which member from Datasharing group11_Mukesh Jain21_jain.mukesh@AOINS.COM31_Tue, 22 Sep 2009 16:53:18 +0000407_utf-8 Hi Sameer, Which version od DB2 you are running. In version 7 and 8 i believe it determine from the order of DB2 subsystem from IEFSSNxx in your parmlib. When you IPL the subsystem are registered in the order defined in IEFSSNxx and when you submit batch job it will scan and try to connect to the first DB2 subsytem defined in IEFSSNxx. If the first one is not available it will try second one. [...] 323 53 39_Re: Which member from Datasharing group20_Cuneyt Goksu (Gmail)22_cuneyt.goksu@GMAIL.COM31_Tue, 22 Sep 2009 20:55:27 +0300580_us-ascii As far as I know depends on the order in member IEFSSNXX...

Regards, Cuneyt

> -----Original Message----- > From: DB2 Data Base Discussion List > [mailto:DB2-L@IDUGDB2-L.ORG] On Behalf Of Sameer Rana > Sent: Tuesday, September 22, 2009 7:15 PM > To: DB2-L@WWW.IDUGDB2-L.ORG > Subject: [DB2-L] Which member from Datasharing group > > > If I submit a batch job on an LPAR A which has two members ( DB3Z and > DB5Z ) of a datasharing group DSNZ and I refer to DSN(DB2G) as the > generic group attach name in my job and SYSAFF LPAR A.What governs > whether [...] 377 77 39_Re: Which member from Datasharing group13_Taddei, Cathy27_Cathy.Taddei@PACIFICORP.COM31_Tue, 22 Sep 2009 10:56:32 -0700425_us-ascii Hi Sameer. It depends on which version of DB2 for z/OS you are running. In version 8 and prior, the job will run in the member that appears first in the subsystem name table (SYS1.PARMLIB(IEFSSNxx). In version 9, it will go to the different members on an LPAR in a round-robin fasion. IBM has accepted my enhancement request to allow it to go back to the old way, but I do not know when it will be implemented. [...] 455 804 28_Re: identity column db2 unix14_Brenda Boshoff21_b.boshoff@UTORONTO.CA31_Tue, 22 Sep 2009 15:13:27 -0400482_iso-8859-1 Hi,

I think you might be looking for SYSCAT.SEQUENCES table, and column:

NEXTCACHEFIRSTVALUE where SEQTYPE = 'I'

That holds, "The first value available to be assigned in the next cache block. If no caching, the next value available to be assigned."

I don't think you will see the actual value that you have restarted the column with anywhere, but you can see what it will use next and figure out if that figure is what you are expecting. [...] 1260 205 28_Re: identity column db2 unix13_Karin Hilbert12_ioh1@PSU.EDU31_Tue, 22 Sep 2009 16:01:49 -0400622_ISO-8859-1 There is a NEXTCACHEFIRSTVALUE in SYSCAT.COLIDENTATTRIBUTES

When I did a select of the nextcachefirstvalue for one of my test tables I got a value of 136, which was the next value to be used. Then I did an "alter table schema.tablename alter column colname restart with 150;" command and then the select of nextcachefirstvalue was 0. I then added a new record in the test table and the identity value for that record was 150. After the record was added, the nextcachefirstvalue returned 151. I hope that helps, Karin On Tue, Sep 22, 2009 at 12:33 PM, Renzo razzetti wrote: [...] 1466 261 28_Re: identity column db2 unix13_Karin Hilbert12_ioh1@PSU.EDU31_Tue, 22 Sep 2009 16:13:50 -0400339_ISO-8859-1 I did the same thing again with the SYSCAT.SEQUENCES table:

The initial select on nextcachefirstvalue showed 151. After the alter statement to alter it to 175, the select showed 0 for nextcachefirstvalue. And after the next record was entered, which used 175 for the id column, the select showed 176.

- Karin [...] 1728 27 49_Modify Cleanup Before Drop Tablespace or Database9_Dan Chatt15_dchatt@FNNI.COM31_Wed, 23 Sep 2009 00:24:52 +0000339_utf-8 Is it still necessary to do a MODIFY RECOVERY and MODIFY STATISTICS before a table space or data base is dropped to clean up the entries in the catalog or will this happen with the DROP? In the old days I think there was an issue other than just performance where it was recommended to do the MODIFY first followed by the DROP. [...] 1756 77 36_Committing from CICS to a remote DB215_Frank Swarbrick30_frank.swarbrick@EFIRSTBANK.COM31_Tue, 22 Sep 2009 18:29:42 -0600329_us-ascii We are a VSE shop that exclusively uses DB2/LUW remote databases.

In VSE when we are done with all of our SQL we just issue EXEC CICS COMMIT RELEASE. This both commits the current LUW and severs the connection to the application server. This is not available on anything other than VSE as far as I can tell. [...] 1834 25 22_Re: OSC and LOB spaces9_Dan Chatt15_dchatt@FNNI.COM31_Wed, 23 Sep 2009 00:36:03 +0000404_utf-8 We did rebinds for all of our developement versioned packages and we are a small shop. The DSN_QUERY_TABLE grew to more than 2GB. I really don't like having catalog objects this large. What will be missing in OSC if we don't create this table?

_____________________________________________________________________

* IDUG 09 Rome, Italy * October 5-9, 2009 * http://IDUG.ORG/EU * [...] 1860 33 53_Re: Modify Cleanup Before Drop Tablespace or Database16_Rasmussen, Steen22_Steen.Rasmussen@CA.COM31_Tue, 22 Sep 2009 21:06:24 -0400309_UTF-8 In order to clean up the catalog, the DROP will take care of that. Another issue is SYSLGRNX which didn't used to be cleaned of during a drop. I must admit I'm not quite on top where DB2 is in terms of taking care of entries in SYSLGRNX.

Steen Rasmussen CA Sr Engineering Services Architect [...] 1894 127 53_Re: Modify Cleanup Before Drop Tablespace or Database0_17_jtonchick@AOL.COM31_Wed, 23 Sep 2009 00:03:35 -0400553_us-ascii

It depends.? Yes the DROP will remove the SYSCOPY and SYSHIST entries, but at a cost.?





The cost is locking.? DB2 has to take locks on all the data rows being deleted in all of the tables in DSNDB06 that contain information related to the object(s).? If you are in a data sharing environment, this could overwhelm the global locking mechinism and cause all active threads in the entire data sharing group to abend.? Even in in a stand alone subsystem, the locking?load could prevent the DROP from completing. [...] 2022 197 22_Re: OSC and LOB spaces11_Roy Boxwell16_R.Boxwell@SEG.DE31_Wed, 23 Sep 2009 07:51:51 +0200411_ISO-8859-1 I dropped them all and I have not noticed anything "not working" in OSC....very odd...I wonder what on earth they are there for....apart from keeping disks full that is....





Roy Boxwell SOFTWARE ENGINEERING GMBH -Product Development- Robert-Stolz-Straße 5 40470 Düsseldorf/Germany Tel. +49 (0)211 96149-675 Fax +49 (0)211 96149-32 Email: R.Boxwell@seg.de http://www.seg.de [...] 2220 286 39_Re: Which member from Datasharing group11_Roy Boxwell16_R.Boxwell@SEG.DE31_Wed, 23 Sep 2009 07:54:57 +0200607_ISO-8859-1 another reason for CAF connect... If you use CAF to do the connect you get back a pointer to the EIB which then contains the data you need

01 EIB-DATA. 05 EIB-CODE PIC S9(04) BINARY. 05 EIB-TLEN PIC S9(04) BINARY. 05 EIB-EYECATCHER PIC X(04). 05 EIB-SSID PIC X(04). * In non-datasharing following fields can be blank 05 EIB-GROUPATTACH PIC X(04). 05 EIB-GROUPNAME PIC X(08). 05 EIB-MEMBERNAME PIC X(08). * DSNX§XAC/SECLABEL class active? Y/N 05 EIB-XSEC PIC X(01). 05 EIB-FLAG PIC X(01). * X'80' - RESTART LIGHT has INDOUBT threads * X'40' - DB2 Running in NFM 05 EIB-RSV1 PIC X(06). [...] 2507 107 39_Re: Which member from Datasharing group12_McLaren Phil32_Phil.McLaren@AXAWINTERTHUR.CO.UK31_Wed, 23 Sep 2009 11:37:03 +0100559_us-ascii Willie has documented enhancements in this area recently...

http://it.toolbox.com/blogs/db2zos/apar-friday-negating-group-attachs-randomization-34050

Dependent on PK79228 and PK79327. Available in V9 since April 2009. Behaviour can be controlled via new ZPARM RANDOMATT=YES/NO.

Phil

-----Original Message----- From: DB2 Data Base Discussion List [mailto:DB2-L@IDUGDB2-L.ORG] On Behalf Of Taddei, Cathy Sent: 22 September 2009 18:57 To: DB2-L@WWW.IDUGDB2-L.ORG Subject: Re: [DB2-L] Which member from Datasharing group [...] 2615 72 53_Re: Modify Cleanup Before Drop Tablespace or Database10_Roger Hecq18_Roger.Hecq@UBS.COM31_Wed, 23 Sep 2009 09:20:39 -0400350_US-ASCII The recommendation was to do the MODIFY before the DROP and to do both as multiple units of work if there were numerous objects involved. The objective was always to reduce locking contention. The DROPs should also be done from the bottom upwards, to reduce contention.

Roger Hecq MF IB USA DB Support 203-719-0492 / 19-337-0492 [...] 2688 91 53_Re: Modify Cleanup Before Drop Tablespace or Database10_Max Scarpa16_mscarpa@CESVE.IT31_Wed, 23 Sep 2009 16:30:56 +0200541_US-ASCII Just a small but significant add-on: save complete DDL and data unloading them on a tape (or VTS or disks) if tables contain important data ($$$$ data, even if obsolete). There's always someone who later could say: 'I real need that data' and if it's your CEO or someone in the heavens you can justify DROP in a zillion of ways but you did a mistake by default. Or, at least, ask for something written from owner of data. Don't take for granted these data will never use in the future, it's always better to wear kevlar pants. [...] 2780 31 58_Re: DB2-UDB fro LUW - restarted identity column next value13_Jack Campbell23_jackrcampbell@YAHOO.COM31_Wed, 23 Sep 2009 16:57:25 +0000695_utf-8 One option - use dblook it will list the DDL for the table including the alter for the identity column with restart value

for example on TEST.EMPLOYEE - db2look returned:

ALTER TABLE "TEST "."EMPLOYEE" ALTER COLUMN "EMPLOYEE_NO" RESTART WITH 14;

Regards

Jack

_____________________________________________________________________

* IDUG 09 Rome, Italy * October 5-9, 2009 * http://IDUG.ORG/EU *

_____________________________________________________________________

IDUG Europe Attendee Testimonial- "Conference was superb! Couldn’t fault anything. Very impressed." _____________________________________________________________________ 2812 130 58_Re: DB2-UDB fro LUW - restarted identity column next value13_Karin Hilbert12_ioh1@PSU.EDU31_Wed, 23 Sep 2009 13:48:22 -0400499_windows-1252 I was going to suggest that also, but I found the same issue as with the NEXTCACHEFIRSTVALUE in the catalog tables - until you enter a new record, the restart statement doesn't show up. I restarted a test table with a restart value of 200, then generated the ddl with db2look. There was now restart statement in the ddl for my test table. I then added a record, which had the ID value of 200. Then when I regenerated the ddl, it showed the restart statement with the value of 201: [...] 2943 186 58_Re: DB2-UDB fro LUW - restarted identity column next value13_Karin Hilbert12_ioh1@PSU.EDU31_Wed, 23 Sep 2009 14:19:36 -0400318_windows-1252 If I need to restart an identity value, I just execute the following stmt to provide me with the restart value:

db2 "select max(EVENT_ID) +1 from DB2INST1.TEST_EVENTS" | tail -4 | head -1 201 because I always want to restart it with the next highest value from the highest value already used. [...] 3130 171 39_Re: Which member from Datasharing group13_Taddei, Cathy27_Cathy.Taddei@PACIFICORP.COM31_Wed, 23 Sep 2009 13:06:40 -0700346_us-ascii Thanks, Phil, that's very helpful. It will be very amusing to see how long it takes my IBM sales rep to give me this information (I was advised to get status on enhancement requests through sales). Can't wait to get back to the old way; it was convenient when introducing changes to know exactly which member would get the traffic. [...] 3302 73 49_db2 luw (question): New database or New Instance?13_Jack Campbell23_jackrcampbell@YAHOO.COM31_Wed, 23 Sep 2009 23:42:29 +0000453_UTF-8 Dear listers

I am pretty new to DB2 LUW, most of my experience is z/os based. And I have a fairly high conceptual level question

The client wants to add a NEW and SIGNIFICANT workload to an existing DB2 LUW v8.2 server. Current system is OLTP, 60+ million update transactions per day (MQ broker), which consumes approx 50% of CPU/memory on the available server (P560). Current system has HACMP (failover) and HADR(DR recovery). [...] 3376 56 35_Rhonda Hester is out of the office.13_Rhonda Hester18_rhester@US.IBM.COM31_Wed, 23 Sep 2009 22:53:17 -0600478_US-ASCII

I will be out of the office starting 09/23/2009 and will not return until 09/28/2009.

I will respond to your message when I return. For State of Georgia DB2 Support pleas contact Gilbert Kelly at kellygi@us.ibm.com or contact the Consolidated Service Desk at 1-877-GTA-3233.

Thank you, Rhonda

_____________________________________________________________________

* IDUG 09 Rome, Italy * October 5-9, 2009 * http://IDUG.ORG/EU * [...] 3433 59 71_Which member from Datasharing group Which member from Datasharing group0_29_Sherri.Sterling@BPD.TREAS.GOV31_Thu, 24 Sep 2009 08:29:15 -0400281_US-ASCII We use the job class for batch jobs. Initiators can only be up on one LPAR at a time and these initiators have certain classes associated with them. You can look at your joblog with SDSF and see where your job ran. It is about the second or third line of the joblog. [...] 3493 63 40_DB2 Command Line Processor in DB2 on zOS14_Ramachandran A25_ramachandran2yk@GMAIL.COM31_Thu, 24 Sep 2009 18:23:59 +0530345_ISO-8859-1 Hi All,

One of our user needs to enable/configure DB2 Command Line Processor(CLP) in DB2 v8.1 on zOS system. But i am not sure this feature is available in DB2 on zOS.

I know DB2 LUW provideds CLP thought which we can execute various administrative commands.

Can anyone pls suggest some inputs on this same. [...] 3557 367 44_Re: DB2 Command Line Processor in DB2 on zOS13_Seibert, Dave26_Dave.Seibert@COMPUWARE.COM31_Thu, 24 Sep 2009 09:27:27 -0400427_utf-8 Hello Ram,





There is no CLP on z/OS.

Please expand on what your user ultimately wants to accomplish.

You can use a CLP from LUW to get to z/OS if you have DB2 Connect.





With DB2 on z/OS, under ISPF there are a number of tools to process SQL or commands.





Please expand and cite an example or 3 to illustrate what your user wants to do. [...] 3925 31 22_Re: OSC and LOB spaces15_Patrick Bossman25_patrick.bossman@GMAIL.COM31_Thu, 24 Sep 2009 14:18:04 +0000496_UTF-8 Hello, With regards to DSN_QUERY_TABLE growth, we're aware of the problem. You should track APAR PK84380.

This table is primarily used within OSC for query annotation.

Optimization Expert uses this table with some of the other advisers (query adviser, and index adviser will also use it).

Best regards, Pat Bossman

_____________________________________________________________________

* IDUG 09 Rome, Italy * October 5-9, 2009 * http://IDUG.ORG/EU * [...] 3957 505 44_Re: DB2 Command Line Processor in DB2 on zOS14_Sevetson, Phil22_PSevetson@FISA.NYC.GOV31_Thu, 24 Sep 2009 10:33:01 -0400364_us-ascii There is a CLP; you have to install the DB2 Client on your workstation and have the ability to connect to a DB2 Connect (Enterprise edition on a sever somewhere, or Personal edititon on your own PC) which has your target subsystem in its database. I'm rusty on the details, as it's been a long time since I set up a DB2 V8 subsystem for this. Sorry. [...] 4463 143 44_Re: DB2 Command Line Processor in DB2 on zOS11_Mukesh Jain21_jain.mukesh@AOINS.COM31_Thu, 24 Sep 2009 14:46:47 +0000338_utf-8 Hi Ram, Are you planning to use DB2 connect or direct connect through JDBC driver. I am not sure whether CLP is available for DB2 v8 or not but it is available in DB2 v9 for z/OS.

Below is the process:

Prior to using Command Line Processor(CLP), the users must follow the following instructions to configure CLP: [...] 4607 172 58_Re: DB2-UDB fro LUW - restarted identity column next value14_renzo.razzetti24_renzo.razzetti@GMAIL.COM31_Thu, 24 Sep 2009 12:06:01 -0300574_windows-1252 So, in conclusion, there is not way in DB2 Unix to watch the RESTART value in the identity column, after do an ALTER with restart .

Thanks



On Wed, Sep 23, 2009 at 2:48 PM, Karin Hilbert wrote:

> I was going to suggest that also, but I found the same issue as with the > NEXTCACHEFIRSTVALUE in the catalog tables - until you enter a new record, > the restart statement doesn't show up. I restarted a test table with a > restart value of 200, then generated the ddl with db2look. There was now > restart statement [...] 4780 39 45_DB2 Z/OS Version 9 Real Time Statistics (RTS)14_Jeffrey Schade15_jschade@ISO.COM31_Thu, 24 Sep 2009 15:01:27 +0000588_utf-8 This past weekend I ran into huge problems trying to REBUILD an index for a base table that contains 7.8 billion rows. Although I had plenty of space in the sort pool DFSORT dynamic allocation was not working properly causing out of space failures during the unload phase most times after an hour of processing. IBM recommended hardcoding the sort works and this got us past the issue. They then recommended turning on UTSORTAL=Y zparm to allow DB2 to use RTS for controlling the dynamic allocations. Is anyone using this option? I have checked and we have the RTS tablespaces [...] 4820 60 44_Re: DB2 Command Line Processor in DB2 on zOS14_Avram Friedman25_ibmsysprog@GEEK-SITES.COM31_Thu, 24 Sep 2009 16:52:02 +0000646_utf-8 Have you looked at the DSN command, It has been part of mainframe DB2 since day 1.

Avram Friedman

On Thu, 24 Sep 2009 18:23:59 +0530, Ramachandran A wrote:

>Hi All, > >One of our user needs to enable/configure DB2 Command Line Processor (CLP) in >DB2 v8.1 on zOS system. But i am not sure this feature is available in DB2 >on zOS. > >I know DB2 LUW provideds CLP thought which we can execute various >administrative commands. > >Can anyone pls suggest some inputs on this same. > > >Thanks & Regards, >Ram. > >_________________________________________________________ ____________ > >* [...] 4881 373 43_Switch Phase for DB2 for z/OS Online Reorgs22_Frank C. Fillmore, Jr.19_tfgpublic@GMAIL.COM31_Thu, 24 Sep 2009 13:05:06 -0400614_us-ascii One of the design points for DB2 for z/OS over the last several versions has been high or near-continuous availability. Right now I am working with a Q Replication customer, but since the Q Apply component is just a DB2 application, this is a generic DB2 question.





At times we have received a -904 "unavailable resource" message from DB2 when the Q Apply program attempted I/U/D operations against a table undergoing an Online Reorg during the switch phase. This causes an outage in Q Replication. There is a new facility in Q Replication which would enable us to divert I/U/D [...] 5255 466 42_DB2 High Availability - Client Perspective22_Frank C. Fillmore, Jr.19_tfgpublic@GMAIL.COM31_Thu, 24 Sep 2009 13:26:59 -0400459_us-ascii Listers,

I am evaluating DB2 high availability infrastructure for a customer. I am very comfortable with IBM's replica solutions from the database server perspective: SQL Replication, Q Replication, and HADR. I am also familiar with xkoto Gridscale (the customer choose not to go in this direction). The questions are focused on the client side of the equation. In this case the clients are ~ 100 WebSphere and JBoss application servers. [...] 5722 190 47_Re: Switch Phase for DB2 for z/OS Online Reorgs10_Dave Nance16_dav1mo@YAHOO.COM31_Thu, 24 Sep 2009 12:45:55 -0700646_iso-8859-1 Hi Frank, Not sure about best practices on this. The irst thing that comes to mind is when you use maxro(defer) when the reorg hits the log phase there is a WTO (Write to Operator) msg about the logging. You could use operator automation software to have that WTO trigger your job that would issue the Q Replication command and altering of the reorg to maxro(5) and then your next Q Replication command. David Nance ________________________________ From: "Frank C. Fillmore, Jr." To: DB2-L@WWW.IDUGDB2-L.ORG Sent: Thursday, September 24, 2009 1:05:06 PM Subject: [DB2-L] Switch Phase for DB2 for z/OS Online [...] 5913 217 39_Re: Which member from Datasharing group17_Dell'Anno, Aurora22_Aurora.Dellanno@CA.COM31_Thu, 24 Sep 2009 21:53:12 +0100593_us-ascii I find it fascinating how going back to old, trusted and sensible behaviour classes as enhancement...







Thanks.





Aurora













Aurora Emanuela Dell'Anno CA Sr. Engineering Services Architect Tel: +44 (0)1753 577 733 Mobile: +44 (0)7768 235 339 Aurora.Dellanno@ca.com











P please don't print this e-mail unless you really need to!













IDUG Europe 2009 - Rome, Italy - October 5-9, 2009 [...] 6131 241 49_Re: DB2 Z/OS Version 9 Real Time Statistics (RTS)16_Christian Michel18_cmichel@DE.IBM.COM31_Thu, 24 Sep 2009 23:39:19 +0200634_ISO-8859-1 We have many customers using that enhancement by now (it has been shipped 1.5 years ago). The TOTALROWS and TOTALENTRIES columns will be populated only by REORG TABLESPACE, LOAD REPLACE and REBUILD INDEX. You're correct that RUNSTATS will not update those columns, but then most of our customers are running RUNSTATS with SHRLEVEL CHANGE which would not be able to capture an exact count of rows/keys that could be written to those RTS tables. But once you activated Real-Time Statistics these columns will populate over time whenever you run maintance (REORG TABLESPACE or REBUILD INDEX) against your objects. We are [...] 6373 604 49_Re: DB2 Z/OS Version 9 Real Time Statistics (RTS)15_Schade, Jeffrey15_JSchade@ISO.COM31_Thu, 24 Sep 2009 18:16:49 -0400489_iso-8859-1 Thank you Christian. I was hoping for RUNSTATS to update the counts since dynamic allocation without RTS (The fall back logic) was not working and to be honest the tables with billions of rows are not routinely maintained with REORG or REBUILD due to the volume of data. I am glad to hear that many customers are using this enhancement. It makes me feel comfortable that once we can get the RTS statistics populated my problems from this past weekend will not be an issue. [...] 6978 478 49_Re: DB2 Z/OS Version 9 Real Time Statistics (RTS)12_Ayalew Kassa20_ayalakassa@YAHOO.COM31_Thu, 24 Sep 2009 20:03:57 -0700426_iso-8859-1 We enabled RTS in 2007 but just changed UTSORTAL ZPARM to Y. When we enabled RTS, we forgot to intialize totalrows and totalentries. Due to this zparm change , we got reorg and load job abends because the total number of rows in the RTS and the real number of rows for DB2 objects was not the same. So, if you have not already intialize the two columns, make sure to set them to NULL before setting ZPARM = Y. [...] 7457 50 39_Re: Which member from Datasharing group14_James Campbell25_jacampbell@ACSLINK.NET.AU31_Fri, 25 Sep 2009 05:13:13 +0000503_utf-8 1) V9? I think GETVARIABLE('SYSIBM.SSID') provides this. In V8 you can modify the authorisation exit to set up a session variable to contain the member name.

2) re comments about batch job classes: some sites set up WLM Scheduling environments - the environment is then started, or stopped, as subsystems become available, or not. Jobs include SCHENV= on the Job card to ensure they only run when and where the subsystem is available, or the type (eg DAY, NITE, WEND) of availability. [...] 7508 475 47_Re: Switch Phase for DB2 for z/OS Online Reorgs13_Bell, Raymond20_Raymond_Bell@BMC.COM31_Fri, 25 Sep 2009 09:46:05 +0200603_us-ascii Hi Frank,

Just a thought, but does this customer of yours have (or has given any thought to acquiring) an ISV Reorg utility? There are some out there that handle the whole Logapply/Logfinal/dataset switching phases much more intelligently. From the utility's perspective, it's possible to control how long to wait for your drain request, how long to wait for a retry and now many times to retry. If we (well, c'mon, whose reorg utility was I gonna be talking about?) still can't get the drain 'cause, say, Q-Rep is busy populating the table then we fail gracefully. Sure, you don't [...] 7984 544 49_Re: DB2 Z/OS Version 9 Real Time Statistics (RTS)16_Christian Michel18_cmichel@DE.IBM.COM31_Fri, 25 Sep 2009 10:14:47 +0200564_ISO-8859-1 Jeff,

if you have a good idea on how many rows there are in these large table spaces (I guess these are partitioned, so you'd need to know these counts by partition), then you can as well set TOTALROWS to these counts. Even though RTS tables are now in the catalog since DB2 9 NFM, these tables are updateable by the user. While we do not want these tables to be updated all the time, for the current exploiters of TOTALROWS/TOTALENTRIES it is ok to seed those values with actual numbers to overcome sort work data set allocation problems - [...] 8529 77 39_Re: Which member from Datasharing group11_Mike Turner19_mike.turner@GMX.COM31_Fri, 25 Sep 2009 10:03:34 +0100592_UTF-8 Hi James

Its GETVARIABLE('SYSIBM.SYSTEM_NAME') and its available in V8.

Regards Mike Turner Email: mike.turner@gmx.com Tel: +44 (0)1565-873702 Web: www.michael-turner.ltd.uk Michael Turner Ltd. Registered in England & Wales No. 1433722. Registered Office: 131 King Street, Knutsford, Cheshire, WA16 6EJ. VAT No. GB 338 4159 44. ----- Original Message ----- From: "James Campbell" Newsgroups: bit.listserv.db2-l To: Sent: Friday, September 25, 2009 6:13 AM Subject: Re: [DB2-L] Which member from Datasharing group [...] 8607 927 49_Re: DB2 Z/OS Version 9 Real Time Statistics (RTS)15_Schade, Jeffrey15_JSchade@ISO.COM31_Fri, 25 Sep 2009 07:27:07 -0400325_iso-8859-1 Hi Ayalew,





Thank you for the real world experience. I appreciate it since I had more than enough abends this weekend until we finally got it right. At this point it looks like the TOTALROWS/TOTALENTRIES are either NULL or have a valid count. So I am hoping this experience won't occur. [...] 9535 1074 49_Re: DB2 Z/OS Version 9 Real Time Statistics (RTS)15_Schade, Jeffrey15_JSchade@ISO.COM31_Fri, 25 Sep 2009 07:30:26 -0400499_iso-8859-1 Hi Christian,





Thank you again. This gives me another option for the larger tables that cannot easily be REORGED or indexes REBUILT. So I should be able to do Runstats on the tables by partition and manually update the TOTALROWS/TOTALENTRIES columns and still provide DFSORT/DB2 the information it needs to do successful dynamic allocations. This would certainly be easier than manually generating the sortwork DDs all the time and should be a one-time effort. [...] 10610 1062 57_SV: [DB2-L] DB2 Z/OS Version 9 Real Time Statistics (RTS)13_Olle Brostrom25_olle.brostrom@SWEDBANK.SE31_Fri, 25 Sep 2009 14:11:32 +0200405_iso-8859-1 Hi RTS users! be aware of the following apars: PK80947 PK91830 PK82367 PK84691

They all adress problems with the realtime statistics tables, some of the apars does not have any solution yet! We have some reorg utility problems caused by negative numbers or incorrect (too large) numbers in RTS tables. We are using RTS heavily to determine all Reorgs and Image Copy to be executed. [...] 11673 196 39_Re: Which member from Datasharing group12_Simon Pawson27_simonp.lists@GOOGLEMAIL.COM31_Fri, 25 Sep 2009 13:41:21 +0100342_ISO-8859-1 Or the current member special register. I use a noddy TEP2 running on multiple systems inserting timestamps and the current member special register into a table to force allocation of new CF structures when I'm creating a new DSG.

Simon

On Fri, Sep 25, 2009 at 10:03 AM, Mike Turner wrote: [...] 11870 60 61_Re: SV: [DB2-L] DB2 Z/OS Version 9 Real Time Statistics (RTS)10_Max Scarpa16_mscarpa@CESVE.IT31_Fri, 25 Sep 2009 14:51:05 +0200422_US-ASCII Even in V7/8 there were similar errors......

Jeff, tablespaces without TOTALROWS had even UPDATESTATSTIME to NULL ? In V7 (when I started RTS) updating this value to a conventional date caused TOTAROWS to be populated.

Max Scarpa

'It has been proven attending IDUG 2009 in Rome eliminates a well know phenomenon in quantum mechanics, pointed-hairy bosses effect (a.k.a DILBERT effect) ' [...] 11931 32 62_AUTO: Renee Wilson is out of the office (returning 29.09.2009)12_Renee Wilson17_TM2110@CH.IBM.COM31_Fri, 25 Sep 2009 14:55:53 +0200412_US-ASCII I am out of the office until 29.09.2009.

I am currently on leave returning Tuesday 29th September. If its urgent please call my mobile on +41 79 686 9798.

Thanks, Renee



Note: This is an automated response to your message "DB2-L Digest - 25 Sep 2009 - Unfinished" sent on 25/9/09 9:04:51.

This is the only notification you will receive while this person is away. [...] 11964 1329 61_Re: SV: [DB2-L] DB2 Z/OS Version 9 Real Time Statistics (RTS)15_Schade, Jeffrey15_JSchade@ISO.COM31_Fri, 25 Sep 2009 09:45:26 -0400595_iso-8859-1 Thanks Olle. I will check into these.





Jeff

_________________________________________ Jeffrey Schade

Systems Consultant, Technology Engineering



ISO 545 Washington Boulevard Jersey City, NJ 07310 Voice: (201) 469-3738

FAX: (201) 748-1500 jschade@iso.com

From: DB2 Data Base Discussion List [mailto:DB2-L@IDUGDB2-L.ORG] On Behalf Of Olle Brostrom Sent: Friday, September 25, 2009 8:12 AM To: DB2-L@WWW.IDUGDB2-L.ORG Subject: [DB2-L] SV: [DB2-L] DB2 Z/OS Version 9 Real Time Statistics (RTS) [...] 13294 320 61_Re: SV: [DB2-L] DB2 Z/OS Version 9 Real Time Statistics (RTS)15_Schade, Jeffrey15_JSchade@ISO.COM31_Fri, 25 Sep 2009 09:49:05 -0400729_us-ascii Hey Max,





It looks like all my rows have valid UPDATESTATSTIME values (No Nulls). If I am reading your message correctly that is what I want, correct?





Jeff

_________________________________________ Jeffrey Schade

Systems Consultant, Technology Engineering



ISO 545 Washington Boulevard Jersey City, NJ 07310 Voice: (201) 469-3738

FAX: (201) 748-1500 jschade@iso.com

From: DB2 Data Base Discussion List [mailto:DB2-L@IDUGDB2-L.ORG] On Behalf Of Max Scarpa Sent: Friday, September 25, 2009 8:51 AM To: DB2-L@WWW.IDUGDB2-L.ORG Subject: Re: [DB2-L] SV: [DB2-L] DB2 Z/OS Version 9 Real Time Statistics (RTS) [...] 13615 229 61_Re: SV: [DB2-L] DB2 Z/OS Version 9 Real Time Statistics (RTS)10_Max Scarpa16_mscarpa@CESVE.IT31_Fri, 25 Sep 2009 16:41:27 +0200555_US-ASCII Yes. But are there tablespaces having UPDATESTATSTIME not null and TOTALROWS null ? It can't be.

If I remember well (it's more than 2 years now) updating UPDATESTATSTIME with a conventional date via SPUFI (for instance 01-01-2009) was enough to start to populate TOTALROWS via RTS.

Max Scarpa









"Schade, Jeffrey" Sent by: DB2 Data Base Discussion List 25/09/2009 16.22 Please respond to DB2 Database Discussion list at IDUG [...] 13845 612 61_Re: SV: [DB2-L] DB2 Z/OS Version 9 Real Time Statistics (RTS)15_Schade, Jeffrey15_JSchade@ISO.COM31_Fri, 25 Sep 2009 11:53:26 -0400406_us-ascii Then yes, I have an issue because all my rows have current (Today) UPDATESTATSTIME but almost all of them have NULL in the counts columns. I was under the impression we need to do a REORG or LOAD utility to get the row counts updated and maintained.





Jeff

_________________________________________ Jeffrey Schade

Systems Consultant, Technology Engineering [...] 14458 496 61_Re: SV: [DB2-L] DB2 Z/OS Version 9 Real Time Statistics (RTS)10_Max Scarpa16_mscarpa@CESVE.IT31_Fri, 25 Sep 2009 18:46:39 +0200608_ISO-8859-1 Uh, my mistake.

I read again my notes: at the very beginning I updated TOTALROWS and UPDATESTATSTIME selecting number of rows from SYSTABLEPART table and using a particular date as wrote before (even if totalrows is a little different it doesn't matter) . In this way I haven't to reorg ALL tablespaces to have totalrows and I know if a tablespace (after a week or two) wasn't reorganized via scheduler selecting rows from RTS table having the conventional date. In this way other columns are updated when RTS are externalized, as described in another post, without waiting for reorg. [...] 14955 939 61_Re: SV: [DB2-L] DB2 Z/OS Version 9 Real Time Statistics (RTS)15_Schade, Jeffrey15_JSchade@ISO.COM31_Fri, 25 Sep 2009 13:05:39 -0400524_iso-8859-1 Thank you Max. I need to do some testing with manually setting the row counts and see what happens.





I appreciate all the responses to my question and I hope everyone has a great weekend.





Jeff

_________________________________________ Jeffrey Schade

Systems Consultant, Technology Engineering



ISO 545 Washington Boulevard Jersey City, NJ 07310 Voice: (201) 469-3738

FAX: (201) 748-1500 jschade@iso.com [...] 15895 1020 52_Oct 15th - Tridex DB2 for LUW and z/OS meeting in NY20_Arettines, Angela M.29_Angela.Arettines@US.NGRID.COM31_Fri, 25 Sep 2009 13:35:28 -0400551_us-ascii

TRIDEXDB2 for LUW and z/OS



Featured Speaker:

Curt Cotner IBM Fellow,VP,CTO Database Servers





Third Quarterly User Group Meetings DON'T JUST KEEP UP - STAY AHEAD!

All Meetings are FREE. You must Pre-register to Attend. Walk-in registration will not be accepted. A photo ID is required for building access. http://tridexdb2.natemurphy.net





Meeting October 15th, 2009

National Grid, One Metrotech Center, Brooklyn NY [...] 16916 40 33_Data Sharing Errors on DB2Connect13_Karen Vogeley21_kvogeley@NEBRASKA.EDU31_Fri, 25 Sep 2009 13:44:15 -0500520_US-ASCII After three of our data sharing databases (v9) on z/OS 1.10 were setup for workload balancing and fail over we've received errors on DB2Connect -

DIA3202C The TCP/IP call "connect" returned an errno="10060"

And the remote applications are running slower. All DB2Connect servers are at V9.1 FP3a and have the connection concentration and pooling active. These have been on the system even before we upgraded to DB2 v 9 and had no problems until the workload balancing and fail over was setup. [...] 16957 37 37_Re: Data Sharing Errors on DB2Connect11_Mukesh Jain21_jain.mukesh@AOINS.COM31_Fri, 25 Sep 2009 19:16:42 +0000369_utf-8 Hi Karen, Are all these 3 DB2 members in the same LPAR or different LPAR ? It looks like you are getting some timeouts and since you are ggeting problem only after doing work load balancing. You should look at your WLM poilicy again. What is the priority given to distributed task? It looks like DB2DIST is not getting enough resources and getting time out. [...] 16995 23 34_Chris Sutter is out of the office.0_26_csutter@LAWSONPRODUCTS.COM31_Fri, 25 Sep 2009 22:17:50 -0500601_US-ASCII I will be out of the office starting 09/25/2009 and will not return until 09/28/2009.

I will respond to your message when I return.

_____________________________________________________________________

* IDUG 09 Rome, Italy * October 5-9, 2009 * http://IDUG.ORG/EU *

_____________________________________________________________________

IDUG Europe Attendee Testimonial- "The clear and high level technical information of some of the sessions makes the conference a value for the money." _____________________________________________________________________ 17019 214 35_CA DB2 User Day at IDUG Europe 200917_Dell'Anno, Aurora22_Aurora.Dellanno@CA.COM31_Sat, 26 Sep 2009 08:48:59 +0100473_iso-8859-1 Hello list!

Are you going to be attending IDUG Europe 2009 - which takes place in Rome from 5th October, next?

On Sunday, 4th October 2009, CA is hosting a traditional half-day seminar to offer an insight into the DB2 world to all their customers attending IDUG Europe 2009.

The agenda features a User Case Study from Volvo IT on 'Managing performance trends using CA Detector', as well as thought-provoking sessions run by CA experts. [...] 17234 288 86_Reminder! Immediate Volunteer Opportunity at IDUG 2009 - Europe: Moderate a Session!!!17_Dell'Anno, Aurora22_Aurora.Dellanno@CA.COM31_Sat, 26 Sep 2009 08:54:49 +0100449_iso-8859-1 And once again, with only about a week to go, a reminder that you can become an IDUG Volunteer - there are still a FEW sessions open for moderating at IDUG Europe 2009!

==============================================

Dear DB2 colleagues,

The IDUG Europe 2009 conference, which will be held in Rome from 5th October, is almost upon us, and here is a reminder that you can participate and become an IDUG Volunteer! [...] 17523 30 32_Re: Getting started as a DB2 DBA12_Rakesh Kumar19_rks.dudes@GMAIL.COM31_Sat, 26 Sep 2009 11:05:36 +0000337_UTF-8 DB2 Express-C(free) is the one stop for all the learning about DB2. And If one puts enough effort in applying the concepts which is in abundance at @developerworks (google Developerworks).....and by downloading Express-C and getting some hands -on It will her her get on board with IBM DB2.

Best Wishes, Rakesh DB2 DBA [...] 17554 128 58_Re: DB2-UDB fro LUW - restarted identity column next value8_Dan Tuck18_dan.tuck@GMAIL.COM31_Sat, 26 Sep 2009 07:47:38 -0500505_windows-1252 I review the previous answers, but did not see this suggestion.

Select c.tbcreator, c.tbname, c.name, s.increment, s.start, s.maxvalue, s.minvalue, s.cycle, s.cache, s.order , s.LASTASSIGNEDVAL from SYSIBM.SYSCOLUMNS as c inner join SYSIBM.SYSDEPENDENCIES as d on d.dname = c.tbname and d.dschema = c.tbcreator inner join SYSIBM.SYSSEQUENCES as s on s.seqname = d.bname and s.seqschema = d.bschema where c.identity = 'Y' and d.dtype = 'T' order by c.tbcreator, c.tbname, c.name ; [...] 17683 45 34_Yong Dong Wu is out of the office.12_Yong Dong Wu15_wuyd@CN.IBM.COM31_Sun, 27 Sep 2009 17:28:41 +0800491_US-ASCII

I will be out of the office starting 2009-09-27 and will not return until 2009-10-08.

I will respond to your message when I return.

_____________________________________________________________________

* IDUG 09 Rome, Italy * October 5-9, 2009 * http://IDUG.ORG/EU *

_____________________________________________________________________

IDUG Europe Attendee Testimonial- "This was definitely the best IDUG conference I have ever attended." 17729 35 25_Re: Views in DB2 for z/os14_Peter Vanroose17_pvanroose@ABIS.BE31_Sun, 27 Sep 2009 21:01:52 +0000513_UTF-8 It's not the fact that it's a VIEW which is queried that makes (part of) the query potentially materialised. The same may happen with subqueries, nested table expressions, or common table expressions on "plain" tables.

Otherwise said: the query SELECT col1, ... FROM my_view ... WHERE ... on a view "my_view" has exactly the same access path as SELECT col1, ... FROM (SELECT .....) my_view ... WHERE ... where the nested expression between the parentheses is the definition of the view my_view. [...] 17765 106 39_Re: Which member from Datasharing group12_Roger Miller19_millerrl@US.IBM.COM31_Mon, 28 Sep 2009 05:37:36 +0000672_utf-8 If you like the old way, then you need APARs PK79228 & PK79327 for a parameter to have group attach like V8. The PTF came out in April 2009. http://www.ibm.com/support/docview.wss? rs=64&context=SSEPEK&q1=pk79228&uid=swg1PK79228

Roger

On Tue, 22 Sep 2009 10:56:32 -0700, Taddei, Cathy wrote:

>Hi Sameer. It depends on which version of DB2 for z/OS you are running. In version 8 and prior, the job will run in the member that appears first in the subsystem name table (SYS1.PARMLIB(IEFSSNxx). In version 9, it will go to the different members on an LPAR in a round-robin fasion. IBM has accepted my enhancement [...] 17872 225 49_Re: DB2 Z/OS Version 9 Real Time Statistics (RTS)11_Roy Boxwell16_R.Boxwell@SEG.DE31_Mon, 28 Sep 2009 07:46:39 +0200270_ISO-8859-1 You must write a "nice" SQL that fishes the data out of the Catalog (IF it is valid) and then updates all of the TP/TS and IX/IP rows in the RTS. It is a one off update and not that difficult...as long as you are aware of all the tricky little bits..... [...] 18098 72 37_Re: Data Sharing Errors on DB2Connect11_Nenad Vidak21_Nenad.Vidak@GMAIL.COM31_Mon, 28 Sep 2009 06:17:47 +0000311_UTF-8 Hello Karen, check your CONDBAT and MAXDBAT as you may hitting one of those due to increased number of connections on the remaining member. My next step would be checking the syslog and TCPIP logs on z. If nothing found there, you could try with the JDBC trace on a DB2Connect server. Regards, Nenad [...] 18171 742 61_Re: SV: [DB2-L] DB2 Z/OS Version 9 Real Time Statistics (RTS)16_Christian Michel18_cmichel@DE.IBM.COM31_Mon, 28 Sep 2009 10:18:21 +0200366_ISO-8859-1 Yes, current maintenance is useful for RTS, also we did have some issues with UTSORTAL=YES operations in very specific cases that are fixed by APARs.

Actually the PK80947 and PK82367 should handle the negative number problems, PK80947 to ignore those values in utilities, PK82367 to prevent those values from happening in DB2 V8 and DB2 9 CM. [...] 18914 46 37_Fallback to DB2 V82 from a V95 on AIX14_jack fernicola20_bowling730@YAHOO.COM31_Mon, 28 Sep 2009 08:44:21 -0700571_iso-8859-1 Has anyone done a fallback from DB2 UDB V95 to V82 on and AIX system that would like to share their experience?

Thanks Jack

_____________________________________________________________________

* IDUG 09 Rome, Italy * October 5-9, 2009 * http://IDUG.ORG/EU *

_____________________________________________________________________

IDUG Europe Attendee Testimonial- “The main reason for my attendance at IDUG - Europe was general training and rekindling enthusiasm. I took away more knowledge and enthusiasm – IDUG did the job!" 18961 36 70_Management Costs Study: Oracle Database 11g vs. IBM DB2 Enterprise 9.511_Bear, Brian23_Brian.Bear@CHARMING.COM31_Mon, 28 Sep 2009 13:07:29 -0400766_us-ascii Management Costs Study: Oracle Database 11g vs. IBM DB2 Enterprise 9.5

http://i.zdnet.com/whitepapers/Oracle_EM_US_EN_WP_11gIBM.pdf

***************************************************************************** The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. If you received this message in error, please contact the sender and delete the material from any computer. ***************************************************************************** [...] 18998 41 74_Re: Management Costs Study: Oracle Database 11g vs. IBM DB2 Enterprise 9.512_Phil Gunning22_pgunning@GUNNINGTS.COM31_Mon, 28 Sep 2009 18:07:46 +0000465_Windows-1252 Interesting but don't think it is worth the paper it is written on. It will be something that nay sayers who are clueless, management in many cases, can use against DB2. Pg ------Original Message------ From: Bear, Brian Sender: DB2 Data Base Discussion List To: DB2-L@WWW.IDUGDB2-L.ORG ReplyTo: DB2 Database Discussion list at IDUG Subject: [DB2-L] Management Costs Study: Oracle Database 11g vs. IBM DB2 Enterprise 9.5 Sent: Sep 28, 2009 1:07 PM [...] 19040 80 74_Re: Management Costs Study: Oracle Database 11g vs. IBM DB2 Enterprise 9.59_Mike Bell21_mbell11a1@VERIZON.NET31_Mon, 28 Sep 2009 13:47:36 -0500453_us-ascii it is interesting that Oracle is still comparing DBA tasking rather than workload performance or even ease of application coding. I guess it still requires the Oracle DBA to be fully trained in unix scripts once you get past the easy stuff you can do in GUI. I do find the idea of a DBA doing a recovery on a multi-terabyte table with a gui and then leaving the recovery manager running on his pc kind of amusing. It is a different world. [...] 19121 231 74_Re: Management Costs Study: Oracle Database 11g vs. IBM DB2 Enterprise 9.513_Brian Stewart23_briankstewart@GMAIL.COM31_Mon, 28 Sep 2009 13:37:39 -0600526_windows-1252 the 360 TB database was being managed from a gui on pc far away, the pc went out in a storm and the ship was lost

On Mon, Sep 28, 2009 at 12:47 PM, Mike Bell wrote:

> it is interesting that Oracle is still comparing DBA tasking rather than > workload performance or even ease of application coding. I guess it still > requires the Oracle DBA to be fully trained in unix scripts once you get > past the easy stuff you can do in GUI. I do find the idea of a DBA doing a > [...] 19353 154 50_DB2 9 for z/OS Performance Features by Dave Beulke10_DB2usa !!!19_db2usa3@HOTMAIL.COM31_Mon, 28 Sep 2009 17:29:54 -0400695_Windows-1252



Hi DB2 user,

Here is a non-profit Blog about DB2 for z/OS (IBM mainframes): http://db2usa.blogspot.com

Last update on Monday, September 28th 2009



Dave Beulke published several posts about DB2 9 for z/OS performance features on his Blog:





- DSNZPARM parameters - ORDER BY and FETCH FIRST n ROWS in a Sub-Select - Skip Uncommitted Inserts zParm - Universal Tablespace - FETCH WITH CONTINUE SQL - DB2 Version 9 for z/OS Logging enhancement - Performance improvement for IBM DB2 utilities - Data Warehousing and Reporting features - Minimizing the number of SQL call - Too Many Performance Features in DB2 Version [...] 19508 397 39_Re: Which member from Datasharing group13_Taddei, Cathy27_Cathy.Taddei@PACIFICORP.COM31_Mon, 28 Sep 2009 14:32:06 -0700387_us-ascii This is all great information, but the original question was how to tell which data sharing member a job ran in when there was more than one member running on an LPAR, and presumably the program did not happen to record which member it ran in. I don't know an easy way to find that information, except by looking at monitors or SMF. If there is a way, I'd love to know it. [...]