you have several choices - the two normal starting points are
1. change all COPY to EXEC SQL INCLUDE - note this doesn't work if the
copy statement is changeing prefix or something in the copy process.
2. use the DB2 co-processor which ties copy process and precompile all together.
--
Mike
Short answer - LIKE %and% might use an index but just starts scanning
at the front of the index and continues until it finds the first row
that matches. So your timeout will depend on how many rows have to be
bypassed not on the number of rows you fetch. All this processing
occurs in DB2 for the
DIGITS gives you a character string representation of a number so you
are getting an implicit conversion from chars to number on the DIGITS
( t1.emp_no ) , 2 , 9 ) = A.emp_no;
Go to the DB2 listserv or the SQL manual on automatic conversion rules.
Mike
On Mon, Mar 26, 2012 at 9:35 AM, Ron
DBRM CCSID's can come from
1. the default DECP from steplib - If you didn't modify the DECP in
SDSNLOAD and the steplib doesn't include SDSNEXIT then you get the IBM
default (usually with a warning message).
2. precompile parms can override the DECP - and yes, some people will
copy a JCL because
Basic DB2 structure -
1. very old DB2 - plans contained DBRM's - BIND PLAN MEMBER(list of
DBRM's) - DBRM dependencies for these DBRM's are stored in SYSPLANDEP
2. new DB2 (last 15 years) - plans contain package list entries - BIND
PLAN PKLIST(list of package names)
No dependendies stored in
The first thing to check is sequence - there are both one-pass and
2-pass logic for precompiler. Some languages you have an option which
one to use. In one-pass logic - If the sql statement is before the
variable is defined, then you get the message.
More expertise is available on DB2-L
Mike
Some other weird things to check.
1. Is the cpu increase in DB2? check the DB2 accounting records for this.
2. Someone mentioned Cache issues - that would require the program be
compiled without RENT to get working storage in the same memory area -
CBT has a program that will extract cobol
just a guess but I expect the program does dynamic SQL. That is the
only way you can get a different access path without a manual rebind.
CPU speed is part of the DB2 optimization so you should expect some
issues when you do rebinds and install new versions of programs.
And there is more
There are 2 kinds of compression.
The outboard kind that takes place in the tape unit is one example.
there is no difference in the z.os cpu time for writing a compressed
tape.
the operating system kind which is always software.
the software compression can be either just software or hardware
On my copies of the cd for z/os 1.6 - it is on disk 1 and called
z/architecture principles of operation bookshelf.
--
Mike
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@bama.ua.edu
the macros for the IFCIS's are in SDSNMACS - there are a several
macros that expand to the majority of the ifcid's.
there are different macros for SMF100 and SMF101 -
the description is in SDSNIVPD(DSNWMSGS)
DSNWMSGS used to be in SDSNSAMP library.but for V8 and V9 they are now
in SDSNIVPD.
there are several possible questions here - a partitioned dataset (the
DB2 name) needs to be expanded. If the dataset is currently less than
4G and is defined to DB2 correctly, DFDSS will create a new copy with
primary allocation = to current allocation and secondary allocation
unchanged with any
let's take some possibilities.
1. the verify succeeds - there are no hanging end of file or paritial
splits or fatal errors. wha happened to the blocks that had not been
written to dasd yet - well, they are gone into the bit bucket in the
sky.
2. the verify succeeds - and operations is rerunning
The differences are EBCDIC vs ASCII, big endian vs little endian, the joys
of comp-3 and the ever present joys of character stream IO vs z/os record
handling. I do not know of any utility that will handle these changes
automatically.
As described by others, it can be done on a pc, but it is by
There isn't one - the fine manual (from V8) says
grouping-expression cannot include any of the following items:
| A correlated column
| A host variable
| A column function
| Any function that is nondeterministic or that is defined to have | an
external action
| A scalar fullselect
| A CASE
The IDUG DB2-L Listserv is only part of your membership in IDUG. The DB2-L
list archives, FAQ, and delivery preferences are at
http://www.idug.org/lsidug under the Listserv tab. While at the site, you
can also access the IDUG Online Learning Center, Tech Library and Code
Place, see the latest
You mean
//DB2UTIL EXEC PGM=DSNUTILB,PARM='DB7G,REPAIR'
//STEPLIB DD DSN=DSN710.SDSNEXIT,DISP=SHR
// DD DSN=DSN710.SDSNLOAD,DISP=SHR
//SYSPRINT DD SYSOUT=*
//UTPRINT DD SYSOUT=*
//SYSIN DD *
DIAGNOSE DISPLAY MEPL
Mike
On Mon, Feb 25, 2008 at 1:28 PM, Paul Peplinski [EMAIL
there is also an IPCS command that will find the equivalent information in a
dump or even on live DB2 subsystem.
Mike
On Mon, Feb 25, 2008 at 2:55 PM, Paul Peplinski [EMAIL PROTECTED]
wrote:
On Mon, 25 Feb 2008 13:43:01 -0600, Mike Bell [EMAIL PROTECTED] wrote:
You mean
//DB2UTIL EXEC PGM
The whole point of multiple write passes is that the physical head does not
write a track at the same exact physical location every time. With the
right equipment ( disk manufacturers and data recovery experts have it), you
can center the read head off to the side of the track, and read a value
I suspect DB2 needs SYSOPER or SYSADM DB2 authorization. This is DB2
security and it is maintained by DB2. Talk to the DB2 install sysprog's.
My memory was console commands came in with userid SYSOPR but there are
other circumstances that the userid is passed.
Mike
On Jan 13, 2008 8:56 PM, Yan
Check the archives - long discussion on FTP of DFDSS DUMP datasets.
The net I remember was
DFDSS DUMP to disk blksize=32760
TERSE
FTP
TERSE unpack
FTP RESTORE
Mike
--
For IBM-MAIN subscribe / signoff / archive access
I can think of 2 posible causes,
you do not have sufficient space in the INDEX component for any new CI's
therefore
1. any new records, have to go to a new CA ie CA split
2. the more CA's you have (ie increase free space for the CA which means the
data is loaded across more CA's) the more CA's
Your choices are limited to
1. set and release base registers around the FETCH
2. use a SQLDA which looks like
EXEC SQL FETCH SELECT-1
USING DESCRIPTOR :SQLDA1
or something but this requires that you build the SQLDA with the addresses
and datatypes of the variables. See the DB2 manuals for
Design counts - always has - always will.
ADABAS got rewritten to 64bit addressability bufferpools with what they call
ADABAS 2006. I don't know how widely implemented it is.
The design target for ADABAS is closer to IMS than DB2 because it supports
multiple occurences for fields in the same
I supported ADABAS for 2 years about 20 years ago. What I remember was the
SVC that modified itself and still running with 24 bit addressability.
Don't get me wrong - I could write amazing programs in Natural in just a
couple of hours but the other restrictions were painfull. Went on an
The SPUFI is checking DSNZPARM and DSNHDECP for the CCSID. The person who
installs DB2 set those.
It is a good time to review all the CCSID's in all the products because DB2
is now converting between CCSID's if they are different.
CCSID conversion is not free.
--
Mike
Wasn't it the F-16 that rebooted when the Israeli's were doing test flights
over the dead sea? didn't have code to support a negative altitude and the
dead sea is about 1500 feet below sea level.
Mike
--
For IBM-MAIN
adding program TEST to TSO authpgm list will work for a sequence like
//TEST EXEC PGM=IKJEFT01,DYNAMNBR=20
//STEPLIB DSN=SIC8.LOAD,DISP=SHR
//FILE01 DD SYSOUT=*
//SYSTSPRT DD SYSOUT=*
//SYSPRINT DD SYSOUT=*
//SYSTSIN DD *
CALL TEST
//*
But I do not know of any way to run APF authorized
and if the thingie can be disassembled, but it has to be wrapped in bubble
wrap and cardboard
Never mind
--
Mike
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the
-923/-924 are connection failures - check for plan names, link DSNHLI
version. Does RC-Update use a different plan name for updates than selects?
The CA support would probably have a check list.
Mike
On 3/27/07, Walter Davies [EMAIL PROTECTED] wrote:
Hi
We are converting our system to a new
Everyone agrees that DB2 is the future but
1. the only way to convert to DB2 is to redesign to DB2. The quickest way
to need a hardware upgrade is to try to use it as a file access method.
2. Why aren't you doing the normal file swap with transaction logging? Full
application transaction
yes - and it has always been that way. You do not have to recompile your
programs just because you have a new DB2 release nor do you have to relink
them. With V8, there is a suggestion that it is best to REBIND them, but
that is still not a requirement.
There are DB2 programs that have not
More Detail -
Old load modules can access DB2 if they have the supporting libraries to
run. Still in the manuals - still supported.
The interface to DB2 is a module called DSNHLI - It hasn't changed for
ages. Now what it does is load the real processing module from steplib,
joblib, or wherever
When my daughter moved into the dorm at a university, I configured
everything before it was connected to the ethernet outlet in the wall. It
took 3 seconds before zone alarm reported the first probe.
--
Mike
--
For IBM-MAIN
Do all your numbers end in 0? there can also be 1C through 9C.
On 12/21/06, Jeffrey Deaver [EMAIL PROTECTED] wrote:
Crude, but I decided to simply edit my file2 and change x'0c' to x'0d' for
0-9 in the sign column.
--
Mike
Depends if you remember when GM bought EDS (long time ago now) - GM had
mulitple data centers on the same physical location because the different
organizations wouldn't share.
Mike
On 11/25/06, Timothy Sipples [EMAIL PROTECTED] wrote:
Google has a link to a cached copy of the Austin-American
I seem to remember that DB2 would not alter add volume if the volume was
already in the allocation. I think the crucial error is not having the DB2
stogroup defined as colume (*) - which will never match an existing volume.
That will force it to the SMS allocation.
Mike
On 9/20/06, Robert
easiest to tell from the generated code but I suspect the cause is the lack
of sign for pic 999. COBOL seems to force the sign to x'Fx' before it adds
or subtracts anything.
77 COUNTERX PIC 999 VALUE 0.
01 BAD-NUMBERPIC 999.
01 BAD-SPACE
something wrong here - my experience was that db2 would start using any sms
volume as soon as it was enabled.
I would suggest open an incident with IBM.
Some things to check
in DB2 the STOGROUP should have volumes (*)
the priquty and secqty for the db2 tablespace and index spaces should be
DB2 listserv moved to the IDUG servers several years ago - someone else may
remember exactly when but it is now the correct address.
or
[EMAIL PROTECTED]
--
Mike
--
For IBM-MAIN subscribe / signoff / archive access
If the source of the data is DB2 and the DB2 is version 8 you can convert
the data to unicode in the unload.
It does require that you modify the SQL and replace every character column
as
CAST (STMT AS CHAR (16) CCSID UNICODE)
You have to set the length field correctly
If you want to convert
one possibility I haven't seen mentioned is are there 2 DCB;s for the same
ddname?
I remember doing this once but other than make sure not to ever do it again
don't remember what the symptoms were
Mike
On 8/30/06, Andy Wood [EMAIL PROTECTED] wrote:
On Wed, 30 Aug 2006 08:34:19 -0500, Mark
LE as used by COBOL and LE as used by PL/1 are wildly different. The
functionality is the same but many of the interfaces and defaults are
different. For examle, I had a problem whtere the old cobol II runtime was
in the joblist - PL/1 ran just fine - COBOL did NOT.
MIke
On 8/28/06, Johnny Luo
The virtual 8304K looks suspicious. normally that means some 24bit code
because there isn't as much usage of EXT (above the line private).
Mike
JESYSMSG tells me:
IEF374I STEP/TESTME/STOP 2006177.1542
CPU0MIN 00.13SEC
SRB0MIN 00.00SEC
VIRT 8304K
SYS 380K
EXT2952K
SYS
I am not an SMF expert either - google brought me to the Cherl Watson smf
ref.pdf
origional poster specified tape dataset and I thought type 17 was scratch
for any non-vsam dataset.
61,62,63,... 66 are all for VSAM.
Mike
--
For
What you want is SPLICE - from the manual
SPLICE - splices together fields from records that
have the same numeric or character field values
(that is, duplicate values), but different information.
Fields from two or more records can be combined to
create an output record. The fields to be spliced
I am also confused - The only interface to 64 bit virtual that I know of is
media manager code for VSAM - as used by IMS or DB2.
All the rest of the access methods support 31bit virtual with 64 bit real
addresses. This means no more thrashing around in Z/os looking for a below
the line real
RLSE doesn't happen unless the dataset has been opened for output. Which
BR14 doesn't do.
dataset allocated with IEFBR14 is exactly like specified in JCL - complete
with no EOF or anything.
Mike
--
For IBM-MAIN subscribe /
At a previous employer - they designed a external time reference that had
some special features - It could adjust the speed of the clock to adjust to
time changes.
DST coming - run clock double time until you get to an hour ahead.
fallback - run clock half speed till you match up.
every so often
The parm you want is RECOVER LOGONLY - This is discussed in DB2
administration guide and the utilities manual.
The short answer is both DF/DSS, FDR and all the products that do volume
backups can easily do volume restores or dataset restores with the one
exception of multi-volume datasets - then
or you could load it to DB2 and use SQL to get the cartesian join (the SQL
people get excited about calling a many to many merge by different names)
select tab1col1, tab1col2, tab2col2 from input1 inner join input2 on
tab1col2 = tab2col1.
The cartesian join will match duplicate keys from either
STROBE does have a DB2 component but you have to pay for it. Once you have
it, it will identify the cost by SQL statement.
DB2 SMF records are a much higher level reporting. They will show IO, cpu by
package (program) but not by sql statement.
You can do the iebeyeball method and watching the
A couple of repro's with delete define for some different ci sizes will give
you enough data points to answer the esential question.
What does it cost to do a VSAM IO from CICS at Ted's shop?
Mike
--
For IBM-MAIN subscribe /
I don't remember any reduction in cpu instructions for VSAM just enhanced
functionality. One of the reasons DB2 uses the media manager interface.
You ddin't mention the transaction rate. Is this one transaction doing
batch workload? or lots of small transactions?
Since the cache hit ratio is
The VSAM buffer logic isn't that smart - one transaction that does scan of
thousands of records will flush the buffers.
If you have the memory and file size is small enough, a user maintained data
table will keep the entire structure in memory.
The other thing you might consider is making the ci
COBOL by itself doesn't usually s0C4 - COBOL in combination with ASM is
different.
Things to check
1. dynamic vs static call
2. ASM module marked as RENT or REUS
3, ASM truly coded as RENT or REUS
4. structure of the parm list passed to ASM
--
Mike
There are four kinds of things to backup
1. DB2 applications
2. Non database application data
3. DB2 system data (catalog and directory) and log files
4. system data.
And you should back them up in that order because
1. DB2 backups are dependent on the information in tape catalog, MVS catalog
and
usually better chance for an answer on DB2 listserv but I lurk here also.
The ability to generate a DBRM is dependent on the status of the information
in SYSSTMT and SYSPACKSTMT.
BIND COPY accross subsystems does not have enough information to rebuild the
DBRM.
REMOTE binds do not have enough
). . It was a
chance to clean up a long standing mess.
Mike
On 3/24/06, Chase, John [EMAIL PROTECTED] wrote:
-Original Message-
From: IBM Mainframe Discussion List On Behalf Of Mike Bell
[ snip ]
The single biggest issue I remember was blksize on loadlibs.
After much discussion, we
Putty is a telnet replacement not capable of doing 3270.
The only TSO that would work would be the old command line stuff - no ISPF
--
Mike
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to
VSAM is VSAM - so DB2 and IMS that is on VSAM can be moved with DF/DSS or
any other tool. someone already said that.
IMS that is on OSAM is not device dependent but the blksize was probably
chosen for 3380 optimum - The DBA's will have to fix it later. If you mark
the old volumes as disabled for
Actually - the restriction on ASM modules is not impossible to fix.
The LE bookshelf has the information on creating LE enabled ASM modules -
CEEENTRY macro is the starting point.
The basic point is that 'normal' ASM conventions break the LE chain so that
the next LE module will abend. A quote
I am sure there are better ways but simple has it's own benefits.
//ST1 EXEC PGM=ADRDSSU,PARM='TYPRUN=NORUN'
and any restore command will identify the available datasets on the dump
tape
Mike
On 3/20/06, Paul Gilmartin [EMAIL PROTECTED] wrote:
OK. I've never done this before.
I have
One question to check is whether DFSORT is opening and closing the output
dataset.
In other words, Is the dataset initialized with a empty block?
I have gotten various very strange abends where I was reading a file that
was allocated but never initialized.
Mike
This jcl and cblopt works with DB2 V7 and enteprise cobol 3.3 and it works
with either region = 0M or region = 8M
I tried the same options and JCL with DB2 V8 and got an abend - didn't
examine it closely (u3002).
The quickest way out is what was already suggested - change to the stand
alone
One basic question - have you used the tape on the 1.6 image before?
If not, there is a whole list of steps to make sure the device type matches
the z/os device type and the tape is online to unix, so the flex instance
can see it. Check the manuals at funsoft and the flex listserv archives.
Mike
Actually hasn't changed if still single volume - from DFSMS manual.forsecqty -
Allocates an amount of space each time the cluster extends,
as a secondary extent. You can use this secondary allocation to
add space for the data or index components of the cluster. A VSAM
data set can be expanded to
The mod name specifies the format used for the display to the screen. If it
is specified, it will over-ride the default.
Usually the first call to IOPCB will specify the modname, second and later
calls will specify the modname only if they want to change the format.
Mike
On 12/6/05, sudhanshu
Let me extend my answer.
Under normal DB2 operations, the DB2 tabelspace and indexspace datasets open
and close as activity against those datasets comes and goes. Normal numbers
would be 10's to 100's datasets open and close per hour. Any catalog setting
can handle that level of activity. DB2
There isn't a direct function but long string from couple of weeks ago about
going from character to hex will work with the logic reversed.
Unless you just want to convert half word or full word to Decimal in which
case
77 number-half pic s9(4) comp. -- this is cobol definition for asm of
DS
There are 2 different issues here -
1. the MVS catalog that contains the definitions for DB2 utilitized datasets
DB2 does not override anything just issues normal OPEN.
2. the actual DB2 datasets that contain table data and indexes. DB2
overrides everything for this datasets because it uses
use this instead
10 WS-UNPK PIC 9(9)V.
10 WS-PACK.
15 WS-PACKN PIC S9(9) COMP-3.
**
* CONVERT X'FA' THRU X'FF' TO CHAR A THRU F *
* TRANSLATE TABLE
And what happens when the filesize is larger than largeint? That is only
2GB for normal pc signed field.
What happens if the file is open and being extended at that point in time -
PC expects complete ownership of any file you reference.
I expect these limitations on the PC - they are not
That can occur when a module is marked rent, runs from an APF library as
auth,
is loaded into update protected memory (because it is marked RENT and came
from APF library)
and then uses a non-RENT macro or does something to modify the running code.
Open a problem with the vendor, and have them
None, RLSE only works with QSAM/BSAM output files and sortwknn files are
never really initialized.
one of the left over reason that sortwk files are never multi-volume
Mike
On 11/10/05, Jim McAlpine [EMAIL PROTECTED] wrote:
We have a long running COBOL batch job that as part of its' processing
I always used s9(4) comp not binary.
don't remember what binary does right now
Mike
On 11/8/05, Craig Kittendorf [EMAIL PROTECTED] wrote:
Hi,
I've done this before but can't get it to work now. I'm trying to pass an
EXEC PARM field to a batch COBOL program. I have the following:
LINKAGE
If you can't make multiple dataset work, your choices are DEDB using fast
path and DB2.
Did you adjust your buffer pools for the increased number of blocks
necessary?
Did you assign free space to the second dataset?
What was the cause of the high response time?
Mike
On 10/6/05, Mehrdad Rastegar
pointer still
exists. If yes, fix it - see IMS manuals.
The abend is because someone has a job that isn't playing by the rules.
Mike Bell
On 9/28/05, Mehrdad Rastegar [EMAIL PROTECTED] wrote:
Hi all,
In IMS V6R1, How a LOST POINTER may be created in a database and how we
can prevent the Abend
The free adove reader V7 now has a search that can scan all pdfs in
your pc. I have all the zos manuals in one directory so can point it
to there fairly easily.
Mike
--
For IBM-MAIN subscribe / signoff / archive access
Choices I see
1. use the DB2 date facilities when the unload tape is created. REORG
with DISCARD processing will purge records with single column date
manipulation.
2. use INREC to build a DFSORT type date field and then use OMIT logic
from Frank that will let you do 1 year. Use OUTREC to remove
Assuming you have backups, DF/DSS can restore just a single track.
that will overlay the EOF. Problem is that the directory entries might
not be correct so you might still be recovering data.
I don't know of any automated tools for this kind of recovery.
The other choice is to build a track of
The quickest wayI know of is to split the file into 2 files one
contain the 'L' and another that don't.
sort both removing duplicates ( are there multiple ''L' records?)
then merge the files removing duplicates and put the 'L' records first.
MIke
dumb way without using merge
1. sort using your control cards but add INCLUDE to get just the 'L'
2, second sort using same control cards with include anything not 'L'
3. concat 'L' data first other second
sort with your control cards and specify OPTIONS EQUALS which will
keep the 'L' s first
One of the companies I worked for had a standard that was required for
their restart procedures. Output datasets were created in iefbr14
step and then referenced as disp=old in the program step. All the
programs were checkpoint restartable but not with IBM checkpoint. The
default CA7/11 restart
DB2 has a command that will bring it to a consistent state without shutdown,
-ARCHIVE LOG MODE(QUIESCE) will attempt to pause all active users
(programs) at a checkpoiint. If successfull, it will switch to a new
active log and spool the old (truncated) log off. Most shops can't
use it because it
Actually, it doesn't require a microscope. All current hard drives
adjust the track location to compensate for expansion of the platter
because the temperature has changed and other little details. What is
required is a diagnostic controller that can adjust the head position
just to the side of
DFSORT manual is not terribly specific but most numeric conversions
are supported by the TO operand - cut from manual The INREC page says
to see the OUTREC pager which says to see the OUTREC page. you can
specify the output length also ie convert from half word to 8 byte
decimal, etc.
p,m,f,to
The origional statement was that a different program calling the
subroutine didn't have the abend. If the different program was also
ASM and didn't clean up modules loaded or was linked as a static call,
then the storage would still be available for the system to close the
ACB. sort of an
Yes, it brings up task manager
which has a command for shutdown on the command list at the top (not
the tabs - above them)
the last option is switch user - win key + L - which is the lock
Mike
On 6/20/05, Edward E. Jaffe [EMAIL PROTECTED] wrote:
R.S. wrote:
Timothy Sipples wrote:
I
You are looking for spanned records - ie RECFM=VBS.
The COBOL manual has the necessary information.
Also note the difference of using WRITE vs WRITE FROM.
WRITE FROM is required for VBS because that is the only way you can
build a bigger than blksize record.
Mike
On 6/6/05, Andy Robertson
comp-5 is just comp with trunc(bin)
from the manual
Note: The TRUNC(BIN) compiler option causes all binary data items |
(USAGE COMP, COMP-4) to be handled as if they were declared with USAGE
| COMP-5.
I don't remember the old cobol 2 compile options - there is an
equivalent option because it
91 matches
Mail list logo
