ORACLE APPLICATIONS ARCHIVES
Topicwise collection of
Postings on Mail Lists
ON
DBA - Db Related Issues
Database tuning
Date: Wed, 21 Jul 1999 17:07:07 -0500
From: Bharat Patel bpatel@DOMINOAMJET.com
Subject: Tuning
Hi guys I have below output from v$waitstat and v$sysstat.
my undo header value is increasing slightly from 10-15 to 55-60.
dirty buffer inspected value is ranging from 10 to 75 at pick time.
I have 12 rollback seg having initial 7M next 7M min 5 max 11 optimal 35M.
Should I have to increase the number of rollback segs.
my extends value is 0 and wraps is about 1-2 for every rollback segs.
shrink and avgshrink value are zero for all rollback segs.
waits are about 1-2 for half of them.
I have db_writer_processes set =3 and I have three processors. My cpu hicks
are about 50-70%.
before increase this values ( less the nos. of CPU * 2 ), I would like to
tune other things.
Any kind of suggestions are welcome.
thanks
Bharat Patel
D.B.A.
bpatel@dominoamjet.com
847-244-2501 ex 1249
Date: Thu, 22 Jul 1999 21:56:42 PDT
From: "venkat subramaniam" venkatdvs@hotmail.com
To: oraapps-dba@cpa.qc.ca
Subject: Re: Tuning
Hi
The ideal way to find out is there any bottleneck is by enabling
time_statistics = true and start the DB and execute UTLBSTAT/ESTAT at peak
intervals and check the report to see any waits available.
Warning : This parameter may cause degration of performance upto 5%
U can also query v$rollstat,v$transaction to really ascertain the
performance
Basically look for sum(writes) in v$rollstat before an operation and after
an operation.
This should be fine
Some tips
Should be minimum 20 extents
ALTER ROLLBACK SEGMENT R01 OFFLINE;
.Sizing should be done W.R.T largest transaction
Another Thumb Rule : SIZE = 10% of Largest Table Size
SYSTEM rollback segment cannot be taken OFFLINE
cheers
Export and Import of Application Database
Date: Mon, 16 Aug 1999 11:26:43 -0400 (EDT)
From: Madhu Sudan Dada dada@csee.usf.edu
To: oraapps-l@cpa.qc.ca
Subject: Exp and Imp of Apps Databse.
We expored whole Apps Database to compressed file using pipe. We are
unable to import and are getting following error message.
IMP-00037: Character set marker unknown
IMP-00021: operating system error - error code (dec 2, hex 0x2)
IMP-00000: Import terminated unsuccessfully
We tried to uncompress the exported file and it has error. This may
be a very easy question but we are stuck.
Thanks,
Madhu Sudan Dada
Separating Tablespaces
Date: Mon, 16 Aug 1999 14:53:03 -0400
From: "Burden, Milley" MBurden@rpschererna.com
To: "'oraapps-l@cpa.qc.ca'" oraapps-l@cpa.qc.ca
Subject: Cor: Seperating tablespaces
We have been live for several years with a few of the Financials modules.
Our current tablespaces are seperated for each fully implemented module with
different tablespaces for the tables and for the indexes. The shared modules
are all together in a tablespace for the tables and one for the shared
indexes. We now want to implement several more modules. As we are doing that
we want to move the tables and indexes for those modules into their own
tablespaces. We know how to do that on the DBA side of things but we would
like to know if there are any things that we need to do within FinApps to
make this happen? Does anyone know of a white paper or anything that lists
all the steps involved?
Applications: 10.7NCA
Platform: RS/6000 AIX 4.2.1
Database: 7.3.4
Modules: GL, AP, AR
TIA,
Milley Burden
Project Manager - Financial Systems
RPSchererNA
(727) 803-2186
mburden@rpschererna.com mailto:mburden@rpschererna.com
Server Partitioning - 10.7 NCA
Date: Wed, 18 Aug 1999 14:33:14 -0700 (PDT)
From: Mark Myszkowski myszko@yahoo.com
To: OraApps-L@cpa.qc.ca
Subject: 10.7 NCA Server Partitioning
Does anyone run 10.7 NCA with 8.0.x db on NT? Has the required server
partitioning affected your performance?
TIA
Mark Myszkowski
Date: Sun, 22 Aug 1999 09:55:31 -0400
From: "Lee McAllister" avernus@iag.net
To: oraapps-l@cpa.qc.ca
Subject: RE: 10.7 Server Partitioning
We are running on Alpha NT server partitioned. I'm not the DBA so I don't
know many of the details, but needless to say having an extra component to
keep up to date certification wise is not fun (patching seems particularly
bad on NT and the fact that it is not a primary tier, doesn't help). But to
be fair I don't think we can trace any of the many problems we have had on
NT could be blamed to server partitioning. HTH
BTW are you alpha or Intel, maybe we could swap war stories, we are having
trouble locating other implementations on NT? Mail me at
mailto:avernus@iag.net if you are interested.
Regards,
Lee McAllister
Lost Disk and Recovery
Date: Thu, 19 Aug 1999 08:05:43 -0400
From: scoughlin@orioncapital.com
To: oraapps-l@cpa.qc.ca, Oracle-L@dbinfo.com
Subject: Lost Disk and Recovery
Okay DBA's....
Here is the scenario:
Have a bad disk in a volume group that is made up of say 4 disks. Need to
swap
out bad disk and add new one. Here is what we did:
1) Shutdown all my db's (say 4 of them) on my HPUX box running 10.20 and
Oracle
7.3.3
2) Take a cold backup of the file systems on this volume group
3) Swap out the bad disk and add the new one.
4) The unix admin redoes everything that Unix admin's do
5) Then he restores the filesystems that were on the bad disk only
6) When we started the databases, 3 of them got the "datafile x needs media
recovery"
Datafile X for each was on the disk that we replaced. Why would I get
this
error?
I simply did a "alter database xxx recover" and all was well. But why
did I
have to?
The files that were restored to the new disk were the same ones that
were
there before
when the database was shutdown. Does Oracle track information in
headers on
the
physical disk?
7) One of the databases had one of its mirrored control files on one of the
disks that
was replaced. I got an error message indicating that control file 1 was
not
in synch
with control file 2 ( the one on the disk we replaced) . Why ?? The
file
that was restored
was the same one that was there before and the database was down the
whole
time?
Love to hear something back on this. I called Oracle and did not get the
response I was
looking for.
Thanks,
Sean
Date: Thu, 19 Aug 1999 10:22:53 -0400
From: "Shreter, Hilary" ShreterH@orau.gov
Subject: RE: Lost Disk and Recovery
Forgive the fuzzy language: when your checkpoints failed to those disks,
that's how the database "knew" that something was wrong. Side effect of not
being able to write the current transaction number to the files is that they
needed recovery/were out of synch.
Database Crash
Date: Tue, 24 Aug 1999 15:36:06 -0600
From: "Juan C. Alvarado" juan_alvarado@ceteco.com
Subject: DBA:Database Crash
Ceteco Information Technology
Hi Folks,
I'm working on Sun 2.6 and RDBMS Oracle 7.3.3.6, I'm having problems
with the database.It crashes after the following error wrote in the
alert file:
Errors in file /home5/oracle/admin/GTBO/udump/gtbo_ora_786.trc:
ORA-07445: exception encountered: core dump [ssdgetcall()+44] [SIGSEGV]
[Address not mapped to object] [308] [] []
ORA-07445: exception encountered: core dump [ssdgetcall()+44] [SIGSEGV]
[Address not mapped to object] [308] [] []
ORA-07445: exception encountered: core dump [npixfc()+516] [SIGSEGV]
[Address not mapped to object] [116] [] []
Mon Aug 23 11:30:34 1999
Error 2068 trapped in 2PC on transaction 7.69.7421. Cleaning up.
Error stack returned to user:
ORA-02068: following severe error from POSC
ORA-03113: end-of-file on communication channel
Any help will be appreciated
Regards
eteco Information Technology
Lic. Juan Carlos Alvarado
Soporte Base de Datos
Ceteco Honduras
P.O.Box 15002
Tegucigalpa, Honduras juan_alvarado@ceteco.com
Tel: +504-235.75.85
Fax: +504-235.74.38
ICQ#: 42484887
Date: Tue, 24 Aug 1999 15:18:36 -0700
From: McRaney Ronald-MCG32315 rmcraney@itexchsrv2.phx.mcd.mot.com
Subject: RE: Database Crash
Juan,
You might try setting:
async_write and use_ism to FALSE in your initGTBO.ora
Good Luck.
Date: Tue, 24 Aug 1999 15:25:22 -0700
From: Karen Blackwell KBlackwell@rockshox.com
Subject: RE: Database Crash
I am on same platform and rdbms. In checking oerr, these messages indicate
you have a serious problem, possible at the OS level. Here are copies of
the message texts:
$ oerr ora 7445
07445, 00000, "exception encountered: core dump [%s] [%s] [%s] [%s]
[%s] [%s]"
// *Cause: An OS exception occurred which should result in the
creation of a
// core file. This is an internal error.
// *Action: Contact your customer support representative.
$ oerr ora 2068
02068, 00000, "following severe error from %s%s"
// *Cause: A severe error (disconnect, fatal Oracle error) received
from
// the indicated database link. See following error text.
// *Action: Contact the remote system administrator.
$ oerr ora 2068
02068, 00000, "following severe error from %s%s"
// *Cause: A severe error (disconnect, fatal Oracle error) received
from
// the indicated database link. See following error text.
// *Action: Contact the remote system administrator.
$ oerr ora 3113
03113, 00000, "end-of-file on communication channel"
// *Cause:
// *Action:
$
Needless to say you need to contact your UNIX System administrator And / Or
your Customer Support Representative at Oracle. I am sorry I cannot give
much more help on this.
Good luck with the issue.
Karen Blackwell
DBA/System Administrator
RockShox, Inc.
408.570-4912 phone
408.435.7468 fax
Date: Wed, 25 Aug 1999 09:35:33 +0530
From: "GopalaKrishnan K" gopalak@wiproge.med.ge.com
Subject: Re: DBA:Database Crash
I have the similar problem in 8.0.5. and Oracle has sent a patch for this
The patch is called as 805110 pack
Contact Oracle Support
Gopal
Date: Wed, 25 Aug 1999 17:01:52 -0600
From: "Juan C. Alvarado" juan_alvarado@ceteco.com
Subject: DBA-Data Base Crash
Ceteco Information Technology
Hi Folks,
I'm working on Sun 2.6 and RDBMS Oracle 7.3.3.6, I'm having problems
with the database.It crashes after the following error wrote in the
trace file:
Wed Aug 25 07:06:04 1999
*** SESSION ID:(1.1) 1999.08.25.07.06.04.000
Exception signal: 11 (SIGSEGV), code: 2 (Invalid permissions for mapped
object), addr: 0x2e2e18, PC: waitpid()+60
Wed Aug 25 07:06:05 1999
ksedmp: internal or fatal error
ORA-07445: exception encountered: core dump [waitpid()+60] [SIGSEGV] =
[Invalid permissions for mapped object] [3026456] [] []
i really need your help people, its production database
Regards
Ceteco Information Technology
Lic. Juan Carlos Alvarado
Date: Wed, 25 Aug 1999 16:39:34 -0700
From: McRaney Ronald-MCG32315 rmcraney@itexchsrv2.phx.mcd.mot.com
Subject: RE: DBA-Data Base Crash
Juan,
Adjusting the shmmax solaris kernel parameter (up
to the limit of physical memory on the machine) may
get around the problem of SGA size reported.
Setting the use_ism init.ora parameter to false (it
defaults to true) should also prevent svrmgr
crashing on startup. There is definitely a problem
with the interaction between solaris 2.6 and Oracle
with respect to intimate shared memory.
Cheers and Good Luck.
Dropping a database - steps for
Date: Wed, 25 Aug 1999 13:32:38 EDT
From: "venkatesh Ramaswamy" venkyra@hotmail.com
Subject: Steps for dropping a database
Hi,
I need to know the steps involved in dropping a database completely.
The Operating system is Sun Solaris 2.5.
Thanks
Date: Wed, 25 Aug 1999 17:03:01 -0700 (PDT)
From: Kevin Gillins kgillins@yahoo.com
Subject: Re: Steps for dropping a database
That one is easy. Just shutdown the database and delete all the
datafiles related to the database. Remove the pfiles and the entry in
the oratab. Update your listener.ora and tnsnames.ora and you have
totally removed your database. If you want to remove the oracle engine
also you should delete the ORACLE_HOME
Kevin
===
Kevin Gillins
Oracle Applications Consulting
WEB PAGE http://www.geocities.com/SiliconValley/4946
Date: Wed, 25 Aug 1999 19:03:27 -0400
From: fushan@gamma.aei.ca
Subject: Re: Steps for dropping a database
Hi,
Dropping a database completely. I would suggest you
shutdown the database and make a cold backup of the database
just in case you need it back for any reason. After the backup is
success. Shutdown database then delete the datafiles. That it!!!
You may also want to "CLEAN UP" some directories that related
to your oracle database like OFA, $ORACLE_HOME, $APPL_TOP
( if you have ) or other custom set up.
It is not hard to do it.
Fu-shan Louisa Leong
Oracle DBA, Oracle Financial Application
Andersen Consulting
Db Parameter Screen - OraApps 10.7 AutoInstall
Date: Sun, 18 Jul 1999 09:45:18 -0400 (EDT)
From: S F sf3010@yahoo.com
Subject: (Database Parameters Screen)
Hi all,
Currently I'm doing a fresh installation of Oracle Apps 10.7 on HP-UX
11 based on 7.3.4.2. The question that I have is related to
AutoInstall. After selecting the products for which we have license, I
moved into the Database Parameters Screen. I just typed the main data
tablespaces and index tablespaces for the products I was interested in.
These tablespace, I had already created in the database according the
Installation Manual. I don't want to create a dummy tablespace for the
rest of the products. So I just left the default value assuming
AutoInstall will check tablespace just for the selected products.
Instead I got a message saying either create those extra tablespaces or
exclude them. Could anyone tell me how do I exclude the tablespaces for
the extra products. Or is it necessary for me to create those extra
dummy tablespaces?
I desperately need urgent help.
Thanks in advance,
Best regards,
SF
sf3010@yahoo.com
Date: Sun, 18 Jul 1999 07:17:02 -0700
From: "Chalmers, Jeff" Jeff.Chalmers@bactc.com
Subject: RE: OraApps 10.7 AutoInstall (Database Parameters Screen)
You have to supply a valid tablespace. You can use the same
tablespace for multiple products. Depending on what you install the
autoinstall process will install some modules in Shared mode. Make sure you
specify unique tablespace names for the products that will be installed
shared.
Date: Sun, 18 Jul 1999 11:38:26 -0400
From: Jay Weinshenker jweinshe@concentric.net
Cc: sf3010@yahoo.com
Create the extra tablespaces... generally there are 2 accepted
solutions
1) Put all the modules you know you won't be using into 2 tablespaces
SHRD for the data tablespaces and SHRDX for the index tablespaces.
2) Make a data and an index tablespace for each anyhow... admittedly
you'll have a total of around 80 or 90 tablespaces to deal with, but it
really isn't that big of deal.. this also makes life easier down the road
if your client/company decides it wants to purchase and install additional
modules...
Also, make the tablespaces large during the install.. you can always
shrink them later.
Date: Sun, 18 Jul 1999 20:56:50 PDT
From: "shankar govindan" shankargovindan@hotmail.com
Subject: RE: OraApps 10.7 AutoInstall (Database Parameters Screen)
Atleast in 10.7 you have no choice but to install all the products, as apps
will have tables linked to these modules, especially with patches.So create
tablespace accordingly for your liscenced modules and dump the others in the
AOL tablespace or create a new tablespace and dump them there like OTHERS.I
think you have to choose MA or something like that and all the modules can
be set to AOL or OTHER tablespace and then edit the required modules field
to their respective tablespace. This way you save time in typing each and
every field.
Regards
Shankar
Date: Sun, 18 Jul 1999 11:57:02 -0400 (EDT)
From: S F sf3010@yahoo.com
Subject: RE: OraApps 10.7 AutoInstall (Database Parameters Screen)
Thanks Jeff, however I have one question, why is it necessary to have
shared products in unique tablespaces?
Best regards,
SF
sf3010@yahoo.com
Date: Sun, 18 Jul 1999 20:59:58 PDT
From: "shankar govindan" shankargovindan@hotmail.com
Subject: RE: OraApps 10.7 AutoInstall (Database Parameters Screen)
No Hard and Fast rules, just makes your life easier in identifying the
modules. Also useful in IMP/EXP of specific modules during development.
Shankar
Date: Mon, 19 Jul 1999 08:20:08 -0700
From: "Chalmers, Jeff" Jeff.Chalmers@bactc.com
Subject: RE: OraApps 10.7 AutoInstall (Database Parameters Screen)
The objects owned by the shared modules will grow over time. Having
unique tablespaces makes it easier to identify space issues and spread IO
usage. At some point in the future you may have a need to fully install a
shared module. In most cases this will require an increase in space usage.
It is far easier to expand an existing tablespace then to shuffle object
into a new tablespace, to meet changing needs.
Jeff
Date: Mon, 19 Jul 1999 02:50:39 -0400 (EDT)
From: S F sf3010@yahoo.com
Subject: RE: OraApps 10.7 AutoInstall (Database Parameters Screen)
Hi,
I just wanted to thank all those on the list who sent suggestions to me
on this topic. I am planning on creating individual tablespaces for all
including the modules which we are not going to use now, so as to be
prepared for any future new modules.
Please feel free to send me your suggestions/clues if you think they
are relevant to this.
Best regards,
SF
sf3010@yahoo.com
One set of Product Files but 2 Conc. Mgrs. possible
dhirajr@calpst.niit.co.in on 06/30/99 08:37:04 AM
cc: (bcc: Mehul Zaveri/EPPCO_DXB/EPPCO)
Subject: Concurrent Manager
Hello guys your help solicited.
We have installed R-11 version 11.0.2. We have two database instances
production and Demo but have installed only one set of product files. Can
we start the internal concurrent manager seperately for the two
instances. If yes does anybody have a copy of the DCP start file. Any
help would be greatly appreciated.
regards
senthil
Date: Wed, 30 Jun 1999 11:25:31 +0400
From: mzaveri@eppco.co.ae
Subject: Re: Concurrent Manager
I don't think this would be possible.
I just went through the $APPL_TOP/admin/$ORACLE_SID/scripts/apcmctl.sh
This file directly calls say FDVL.env (FDVL is a development instance
11.0.3).
In that case you are required to copy and edit this file for other
database.
Ideal no. of concurrent processes to have
Date: Mon, 19 Jul 1999 11:40:22 +0530
From: "GopalaKrishnan K" gopalak@wiproge.med.ge.com
Subject: Number of conc processes in R11
Hi All !
Here we have a strange problem !
After applying the so called MRC mini pack and
a series of patches my CPU usage zooms to 100 %
when i start the concurrent manager.
I had purged all my concurrent requests which is more than
two days old and and still no use.
I called Oracle support and registerd a TAR for this,
They asked me to reduce the number of standard manager
processes to three and still no use.it is in60 - 80 %
Any solutions / sugesttions.
Before this pacth i used to run the processes without any problem
I don't see any problem in my database
Regards,
K Gopalakrishnan
Database Administrator
GE Medical Systems Limited,
X Ray (South Asia) Pune-INDIA
Pin:411014 Ph # +91 20 6688693 /4=20
Date: Mon, 19 Jul 1999 21:50:17 PDT
From: "shankar govindan" shankargovindan@hotmail.com
Subject: Re: Number of conc processes in R11
Gopal,
True by reducing the STD conc processes the possibilities of the
CPU usage might come down. This is because the Memory is full with so many
processes that the CPU just proceses it and dumps it back creating a clog
and your requests wait for IO.
Why dont you accesses your memory requirement correctly. You could use the
help of oracle since the Table for calculating the same does not come with
the ver 10.7 and above.
If you are on NT then go to TASK manager and see the Processes utilization
of memory.
Also you could try a cold boot of your system, may be you have run away
processes eating at you memory and CPU.
In unix try ps -ef | grep Concurrent_Processes
Shankar
Date: Tue, 20 Jul 1999 10:53:37 +0530
From: Gopalakrishnan K gopalak@wiproge.med.ge.com
Subject: RE: Number of conc processes in R11
Dear Shankar !
Thanks for the detailed update.
After reducing the concurrent processes to
4 still my CPU is heavily utilized , ( ~80 %)
I had registered aTAR with Oracle local support and
Worldwide support. I am waiting for their call.
Regards !
Gopalakrishnan K
Database Administrator
GE Medical Systems Limited.
Pune 411014, INDIA
Date: Tue, 20 Jul 1999 20:53:34 PDT
From: "shankar govindan" shankargovindan@hotmail.com
Subject: RE: Number of conc processes in R11
Gopal,
Just yesterday in the Oracle Support meeting review I had taken up
this issue. For your info : If you reduce the STD conc processes to less
than three then drastic problems comes up. From one of the best support
guys.
What operating system are you in. Probably I can help you tune.
You see gopal , if your memory is eaten up by the concurrent processes and
the clients connecting to your system then there is swapping and paging
happening right. IO takes more CPU time than anything else. Correct me if I
am wrong. Next comes your SQL 's. You have to isolate slowly and check the
performance to zero on to your problem.
Shankar
Date: Wed, 21 Jul 1999 15:44:38 +0530
From: "GopalaKrishnan K" gopalak@wiproge.med.ge.com
Subject: Re: Number of conc processes in R11
Hi Shahkar !
Thank you for your suggestion and support.
Now the problem is solved. Actually one BOM
interace Program went to an infinite loop and it
took the entire CPU.
I am using sun solaris and help in tuning the OS
will be highly appreciated.
You are right I/O takes more CPU than anything else only
when the system is normal. In my case also it was.
All these problems would have resolved by cold booting
the system . Since it is Unix i did not want to recommend
this to my SysAdmin. In this case unix also behaves like
windoze.
Regards,
K Gopalakrishnan
Database Administrator
GE Medical Systems Limited,
X Ray (South Asia) Pune-INDIA
Pin:411014 Ph # +91 20 6688693 /4
End of Communication Channel
Date: Fri, 23 Jul 1999 09:20:36 -0400
From: "Timothy Brewer" tbrewer@tbsolns.com
Subject: End of communication channel... ???
HP-UX 11.0
RDBMS 7.3.4.4
Apps 10.7.0.16.1
Getting an End of FIle on communication channel after conversion from
HP-UX 10.20 to HP-UX 11.0. Further message states Error in reading
DBMS_APPLICATION_INFO.
1. Shut listeners down and ran the program local. Still got the error.
2. Ensured DBMS_APPLICATION_INFO is valid (and generated properly
through adprepdb.sql). All OK.
Anyone seen this?
Timothy J. Brewer
Senior Oracle Consultant
TB Solutions, Inc.
Date: Fri, 23 Jul 1999 10:29:50 -0600
From: "Rodney Womack" rwomack@mho.net
Subject: Re: End of communication channel... ???
I vaguely remember something of that nature. I can't remember the cause but
check your alert log for this message:
Rolling back half complete log switch of thread 1
LGWR: terminating instance due to error 19502
Instance terminated by LGWR, pid = 1593
We were running HPUX 11.0 Oracle Apps 11.02 and I believe the error I was
getting was an "end of communications" error then I got the above detail
from the alert log. Let me know as I have a potential fix.
Rodney
Date: Fri, 23 Jul 1999 21:25:34 PDT
From: "shankar govindan" shankargovindan@hotmail.com
To: oraapps-dba@cpa.qc.ca
Subject: Re: End of communication channel... ???
DBMS_APPLICATION_INFO takes effect after you run pupbld.sql, is'nt it,Or am
I wrong.
Why dont you reinstall SQLNET and try. Also regenerate your APPLSYS.env from
adadmin.
Shankar
Date: Sat, 24 Jul 1999 08:47:23 +0400
From: mzaveri@eppco.co.ae
To: oraapps-dba@cpa.qc.ca
Subject: Re: End of communication channel... ???
One of the reason, why this appeared to me was , updating a table directly,
which was already open as a cursor and was being read (I think write too).
The very common mistake is not to use cursor name again in update statement
and just use rowid. This I saw in many prog's code.
SGA size - when system is shut
Date: Sat, 24 Jul 1999 23:37:27 +0500
From: S V MANSATA kunal@blr.vsnl.net.in
To: oraapps-dba@cpa.qc.ca
Subject: size of sga when system is shut
hi all
i would like to know the size of the SGA , when the system is shut,on my
test database i got the following, please help me out.
i am calculating it according to the following formula
SGA = db_block_size * db_block_buffers
+
shared_pool_size + sort_area_size
+
redo log buffers
when i put show sga at the svrmgrl prompt it gave me the following
total sga 3042576
fixed 48400
variable size 2473984
database buffers 409600
redolog buffers 110592
what i have in my init file is as follows
db_block_size =2048
db_block_buffers=200
shared_pool_size=1750000
sort_area_size=65536
redo_buffers=102400
in addition it adds 8192 bytes to the redolog as u can see
( 110592 -102400)=8192,
why does this happen, can someone show me the correct method
thanking you
promeet s mansata