| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
 |
 | |||
| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
|
| |||
Home -> Community -> Usenet -> c.d.o.server -> Re: Multiple instances on a single database
It's possible to run an OPS configuration on ONE node on linux starting from
Oracle 8.1.7.0.1.
I know it's silly but I'm doing it aswell at home - just as a test set-up to
keep my OPS knowledge alive...
Here's the doc taken from Metalink on how to do it ! Koert
PURPOSE
This document is intended for peoples who would experience Oracle Parallel Server (OPS) technology on a common hardware . This is possible thanks to the OPS options that is newly introduced in the Linux (x86) port of the oracle RDBMS v. 8.1.7.0.1 . The following contents are strictly related to Linux Intel 32bit platform.
SCOPE & APPLICATION
TITLE
Setting up Oracle Parallel Server environment on Linux - Single node
CONTENTS
1.1) DLM and Parallel Cache:
Briefly, Parallel Cache 8i (PC) synchronization relies on a Distributed Lock Management (DLM) software layer that allows lock of resources among all the instances partecipating to the parallel server in a distributed fashion. This layer is part of the RDBMS kernel and mantains an interinstance consistent view of the database of all the locks. DLM is responsable of granting access to protected resources to the instances processes requesting for and tracking the ownership of each lock .
Most of these tasks are carried out by OPS specific background processes :
LMON - The "Lock Monitor Process" monitors the entire cluster to manage
global locks and resources. LMON manages instance and process
deaths and the associated recovery for the
Distributed Lock Manager.
LMD - The "Lock Manager Daemon" is the lock agent process that manages
lock manager service requests for Parallel Cache Management
locks to control access to global locks and resources.
The LMDN process also handles deadlock detection and remote
lock requests. Remote lock requests are requests originating
from another instance
LCK - The "Lock Process" manages non-Parallel Cache Management locking
to coordinate shared resources and remote requests for those
resources
BSP - The "Block Server Process" rolls back uncommitted transactions
for blocks requested by cross-instance read/write requests
and sends the consistent read block to the requestor
1.2) Cluster Management layer :
Cluster Manager (CM) is responsible for maintaining the process-level
cluster status. Cluster Manager accepts registration of Oracle instances
to the cluster and provides a consistent view of which Oracle instances
are alive and which are dead.
Cluster Manager also propagates status information to all the Oracle
instances,enabling communication among instances .
Typically this layer is an Operating System service and depends by the
specific port (Solaris/Aix/HP-UX/OpenVMS), starting with Oracle 8.1.7.0.1
for Linux x86, Oracle Corporation developed a proprietary CM software
that natively interfaces with the 8i kernel (currently it stands out of
the kernel).
DLM relies on CM services to detect cluster events like :
'instance joining the parallel server' or
'process died unexpectedly'
...
1.3) Node Monitor services :
Node Monitor (NM) is responsible for maintaining a consistent view of the nodes, and reports to Cluster Manager which node is alive or dead. The Node Monitors on all nodes in a cluster regularly send ping messages to each other. Each node maintains a database containing status information on the other nodes. The Node Monitors in a cluster mark a node inactive if the node fails to send out a ping message within a defined time interval plus a margin. The Node Monitor of a node can fail to send a ping message due to:
+ the death of Node Monitor?s sending thread on the remote machine + a network failure + abnormally heavy conditions on the remote machine
CM relies on NM services to detect cluster events like :
'network failures' or
'cluster node died unexpectely'
...
1.4) Watchdog (WDT) services [6],[7]
The watchdog device allows applications to make use of a timer facility. First the application should register an action routine with the watchdog device, and start the watchdog running. After this the application must call a watchdog reset function at regular intervals, or else the device will cause the installed action routines to be invoked. The assumption is that the watchdog timer should never trigger unless there has been a serious fault in either the hardware or the software, and the application's action routine should perform an appropriate reset operation.
The main reason for having a watchdog timer is to guard against software and certain hardware failures that may cause the system to stop working, but can be cured by a reset or some other form of restart .
The watchdog device is either implemented by interfacing to a hardware watchdog device, or it is emulated using an alarm on the kernel real-time clock.
'Watchdogd' daemon supplies Watchdog services to Node Monitor, Cluster Manager, and the LMON (Lock Monitor) process by using a Watchdog device. If Watchdogd finds a process that did not notify it during the defined interval, Watchdogd crashes the entire system (!) . On Linux, a Watchdog device is implemented at the kernel level .
NM,CM,DLM rely on WDT services to detect abnormal key processes abends or hangs:
2) Architecture
2.1) Two-node / Two-instance Oracle Parallel Server Archtecture
Shared Disks,
Database on RAW
partitions
/ \
/ \
Node 1 Node 2
--------------- ---------------
| SGA Instance A| |SGA Instance B |
|---------------| |---------------|
| |pmon|__ | | |pmon|___ |
| |smon |lgwr|| | |smon |lgwr||
| ---|dbwr| | | ---|dbwr| |
| ---- | | ---- |
| ------------| |------------ |
| |lmon lmd0 | |lmon lmd0 | |
lipc | | lck0 | nipc | lck0 | | lipc
------- DLM <===================> DLM -------
| | |------------| |------------| | |
| | | CM | | CM | | |
+------> <===================> <------+
| | |___________ | | ___________| | |
| | | NM | | NM | | |
+------> <===================> <------+
| | |___________ | |____________| | |
------> WDT | | WDT <------
| | | | | |
--------------- ---------------
|////kernel/////| |////kernel/////|
--------------- ---------------
'-' lipc : local inter-process communication '=' nipc : networked inter-process communication (via tcp/ip)
Note : Typically the shared disks box is wired via SCSI or Optical Fiber to the main boards, this introduce compatibility issues related to the drivers used and to the controllers used . Please refer to HW certification matrices for that.
2.2) Single-Node / Two-instance Oracle Parallel Server Architecture
Local Disk,
Database on RAW
partitions
|
single Node |
---------------------------------------------
| SGA Instance A| |SGA Instance B |
|--------------- ---------------|
| |pmon|__ |pmon|___ |
| |smon |lgwr| |smon |lgwr||
| ---|dbwr| ---|dbwr| |
| ---- (localhost) ---- |
| __________ nipc __________ |
| | DLM | <=========> | DLM | |
| | lmon | | lmon | |
| | lmd0 lck0| | lmd0 lck0| |
| |__________| |__________| |
| ____|________________________|_____ |
| | | | | |
| | +--> CM svcs <--+ | |
| |____|________________________|_____| |
| | | | | |
| | +--> NM svcs <--+ | |
| |____|________________________|_____| |
| | | | | |
| | --> WDT svcs <-- lipc | |
---------------------------------------------
|//////////////// kernel ////////////////////|
---------------------------------------------
'-' lipc : local inter-process communication '=' nipc : networked inter-process communication (via tcp/ip)
Note : this kind of configuration is not so exposed to compatibility issues related to various SCSI controller and drivers that may limit the implementations .
3) Kernel and System Configuration
Now that the overall architecture is defined, we can start to detail step-by-step the phases that will bring us to the final configuration
3.1) - Bill of Materials [4]
Intel x86 X-windowed and networked box
Red Hat 6.2 distribution
Related Linux kernel source rpms (kernel-source-2.2.14-5.0.i386.rpm)
Oracle 8.1.7.0.1 for Linux Intel
Operating System Linux : kernel 2.2
Operating System Patches: Patch 14 or later
Raw Device Patch Required
System Libraries : GNU C Tools egcs-1.1.2
Disk Space Requirements : 766 MB is typical.
Note: 600KB for Oracle Cluster Manager and
765MB for Oracle8i Enterprise Edition
3.2) - Prepare your disk for RAW partitions
Raw Devices (RD):
OPS on linux needs direct access to datafiles without the interaction with the file system . In fact the Unix file system cache will produce unpredictable effects and potentially corrupt the database due to the fact that blocks that Oracle thinks written on the disk are in effect cached in the file system cache. This normally is not a problem in a non-OPS environment, but when multiple writers have to synchronize on the disk it is necessary that 'unsynchronized' caches are turned-off .
RD feature is available with linux kernels and allow processes direct disk partitions access, without any FS interaction nor cache . To determine if your kernel source tree is patched for Raw Devices simply check if
/usr/src/linux/drivers/char/raw.c exists
Every oracle database datafile,controlfile,rlogfile must be placed on RD so you will need a lot of them . Every RD is bound to a disk partition with 1-1 relation so we will need at least nine, considering a minimal reasonable configuration of two OPS instances for one database :
Raw Devices assignements :
Usage | Object | raw device | Partition | size (MB)
Note : 'OPS_A' and 'OPS_B' will be the instance names )
Note : raw1 is oversized, the right dimension should be at least
4 + [(the number of nodes in the cluster) * 4] KB . [1]
So we will need at least 9 disk partitions on our hard drive . The number of PRIMARY PARTITIONS that you can create on your disk (IDE) is limited to 4 and considering that you are using typically one for root file system, one for swap and one for data you need 8 more . The solution adopted here is to use one EXTENDED PARTITION that allow us to create inside it as many LOGICAL PARTITIONS we need . If your current hard drive is does not containt an EXTENDED PARTITION with at least 500MB free, you will need to repartition it (...) .
Disk Partioning Table : [5]
Disk /dev/hda: 255 heads, 63 sectors, 1467 cylinders Units = cylinders of 16065 * 512 bytes
Device Boot Start End Blocks Id System
/dev/hda1 410 1392 7895947+ 5 Extended <-- Note
/dev/hda2 * 37 256 1767150 83 Linux
/dev/hda3 257 409 1228972+ 6 FAT16
/dev/hda4 1 36 289138+ 84 OS/2
/dev/hda5 410 505 771088+ 6 FAT16
/dev/hda6 506 538 265041 82 Linux swap
/dev/hda7 539 921 3076416 83 Linux
/dev/hda8 922 1334 3317391 83 Linux
/dev/hda9 1335 1335 8001 83 Linux <-- Our LOG.
/dev/hda10 1336 1378 345366 83 Linux PART. are
/dev/hda11 1379 1380 16033+ 83 Linux placed
/dev/hda12 1381 1382 16033+ 83 Linux in the
/dev/hda13 1383 1384 16033+ 83 Linux EXTENDED
/dev/hda14 1385 1386 16033+ 83 Linux Look the
/dev/hda15 1387 1388 16033+ 83 Linux start/end
/dev/hda16 1389 1390 16033+ 83 Linux points
/dev/hda17 1391 1392 16033+ 83 Linux <--
Note : normally after defining the new partions, you'll need to reboot your system in order to re-read the partition table .
Note : RH62 pre-defines in /dev hda1 -> hda16 probably you'will need to add more, use for example '# mknod /dev/hda17 b 3 17 ' .
Now you can bind the newly created partitions to respective raw devices. This binding is volatile and must be restored after any reboot (put in your /etc/rc.local) .
#to be added in system /etc/rc.local
#(part 1/3)
#-- rawdevice binding
/usr/bin/raw /dev/raw/raw1 /dev/hda9
/usr/bin/raw /dev/raw/raw2 /dev/hda10
/usr/bin/raw /dev/raw/raw3 /dev/hda11
/usr/bin/raw /dev/raw/raw4 /dev/hda12
/usr/bin/raw /dev/raw/raw5 /dev/hda13
/usr/bin/raw /dev/raw/raw6 /dev/hda14
/usr/bin/raw /dev/raw/raw7 /dev/hda15
/usr/bin/raw /dev/raw/raw8 /dev/hda16
/usr/bin/raw /dev/raw/raw9 /dev/hda17
#-- print out the current bindings
#
/usr/bin/raw -qa
#-- rawdevice ownership & access rights
# (oracle:dba) is our oracle account
/bin/chmod 600 /dev/raw/raw1
/bin/chmod 600 /dev/raw/raw2
/bin/chmod 600 /dev/raw/raw3
/bin/chmod 600 /dev/raw/raw4
/bin/chmod 600 /dev/raw/raw5
/bin/chmod 600 /dev/raw/raw6
/bin/chmod 600 /dev/raw/raw7
/bin/chmod 600 /dev/raw/raw8
/bin/chmod 600 /dev/raw/raw9
/bin/chown oracle /dev/raw/raw1
/bin/chown oracle /dev/raw/raw2
/bin/chown oracle /dev/raw/raw3
/bin/chown oracle /dev/raw/raw4
/bin/chown oracle /dev/raw/raw5
/bin/chown oracle /dev/raw/raw6
/bin/chown oracle /dev/raw/raw7
/bin/chown oracle /dev/raw/raw8
/bin/chown oracle /dev/raw/raw9
#
# END of (part 1/3)
3.3) - Prepare your kernel for OPS [1]
3.3.1) Kernel sources : check and install if needed the current kernel sources packages .
% rpm -qa | grep kernel
...
kernel-source-2.2.14-5.0.i386.rpm
kernel-headers-2.2.14-5.0.i386.rpm
...
3.3.2) Raw device patch
% ls -l /usr/src/linux/drivers/char/raw.c
3.3.3) Raw device compatible fileutils (optional) : When you use the fileutils package and its dd command to back up raw devices, ensure that you use version 4.0j or later. Earlier versions of fileutils do not support the dd command for raw devices.
fileutils-4.0-21.i386.rpm RPM package
3.3.4) Kernel setup (we assume working on a clean RH62 and you are logged in with a graphic X-session )
% cd /usr/src/linux
% cp configs/kernel-2.2.14-i686.config .config
(we assume a PIII single processor)
% make xconfig
in section 'Watchdog Cards' select to :
'Software Watchdog' = 'm'
'Disable watchdog shutdown on close' = 'y'
return to the main form and select 'Save and Exit'
3.3.5) Kernel & Modules rebuild (experience with linux kernel rebuilds is welcomed as well all the file backups you need before proceeding). We suggest to save the current copy of the kernel and system maps and make a emergency boot disk 'see mkbootdisk' ... [5]
Clean up the old version files of modules by removing files from
/usr/src/linux/include/linux/modules by entering:
% cd /usr/src/linux/include/linux/modules
% rm -f *
Check dependencies. The makefile performs all dependency-checking
procedures after you enter:
% cd /usr/src/linux
% make dep
Clean objects for safety by entering at /usr/src/linux:
% cd /usr/src/linux
% make clean
Make a boot image by entering the following at /usr/src/linux:
% cd /usr/src/linux
% make bzImage
Make sure the boot images finishes without error.
Make modules by entering the following at /usr/src/linux:
% make modules
Install the modules by entering the following at /usr/src/linux,
as root user:
# make modules_install
Copy the bzImage from /usr/src/linux/arch/i386/boot to /boot.
(we overwrote the current image )
# cd /usr/src/linux/arch/i386/boot
# cp bzImage /boot/vmlinuz-2.2.14-5.0
Install System.map file by copying the System.map generated at
/usr/src/linux to /boot.
# cd /usr/src/linux
# cp System.map /boot/System.map-2.2.14-5.0
setup & run 'lilo' to update the boot image with the new
kernel (check linux documentation)
Create the intial ram disk image for the kernel
# cd /usr/src/linux
# /sbin/mkinitrd /boot/initrd-2.2.14-5.0.img 2.2.14-5.0
Edit /etc/lilo.conf by copying a Linux entry for an already-working
kernel. Modify the image and label it initrd for the new entry.
For example :
image=/boot/vmlinuz-2.2.14-5.0
label=new_kernel
initrd=/boot/initrd-2.2.14-5.0.img
read-only
root=/dev/hda2
# /sbin/lilo
Now the new kernel should be registered and ready to boot.
Note that you may build a new kernel and its modules instead of
removing the old ones by simply editing /usr/src/linux/Makefile
and adding an uniquely identified tag to the EXTRAVERSION line
before you begin the build process.
4) install Oracle software
5) Setup Oracle Cluster Manager components
WDT services : Now that kernel supports watchdog devices, we can start the Watchdog Daemon (WD) . After that, the others key process (NM,CM,LMON) of the Oracle Cluster Manager will be able to register with 'watchdogd' and work. WD is responsabile to REBOOT IMMEDIATELY YOUR SYSTEM if some component of Oracle Cluster Manager 'forget' to 'ping' it periodically. The system crash is a solution that prevents that one instance that loses syncronization with the others can survive and continue to operate on the database.
Currently whole system is rebooted but probably in the future lesser drastic solution will be implemented in order to limit the shutdown to the Oracle instances only , like in the other major OPS implementations .
5.1) load the softdog module in your kernel
#to be added in system /etc/rc.local
#(part 2/3)
/sbin/insmod softdog soft_margin=60
#
# END of (part 2/3)
5.2) Create watchdog device
% mknod /dev/watchdog c 10 130
% chmod 600 /dev/watchdog
% chown oracle /dev/watchdog
5.3) Configure the NM
create '$ORACLE_HOME/oracm/admin/nmcfg.ora' and insert the lines :
DefinedNodes=<your node>
CmDiskFile=/dev/raw/raw1
Note : the node name should be returned by 'hostname' and resolved in /etc/hosts (if you're using DHCP you may set the /etc/hosts alias for 127.0.0.1 to the output of 'hostname' )
5.4) NM and CM
You need to define part of the oracle environment in orderd to NM and CM
work . They will generate some further threads (this actually is done
due to a limit in the watchdog services that works on a per-process
basis)
#to be added in system /etc/rc.local
#(part 3/3)
ORACLE_BASE=/u01/ora
ORACLE_HOME=/u01/ora/product/81701
PATH=$PATH:$ORACLE_HOME/bin:$ORACLE_HOME/oracm/bin
LD_LIBRARY_PATH=$ORACLE_HOME/lib:$ORACLE_HOME/oracm/lib
export ORACLE_BASE
export ORACLE_HOME
export PATH
export LD_LIBRARY_PATH
echo "..starting Watchdog daemon"
watchdogd -g dba -e $ORACLE_HOME/oracm/log/watchdogd.log
echo "..starting Oracle Node Monitor"
oranm /m /e:$ORACLE_HOME/oracm/log/nm.log \
</dev/null>$ORACLE_HOME/oracm/log/nm.out 2>&1 &
echo "..starting Oracle Cluster Manager"
oracm /e:$ORACLE_HOME/oracm/log/cm.log \
</dev/null>$ORACLE_HOME/oracm/log/cm.out 2>&1 &
#
# END of (part 3/3)
5.5) Check the /etc/rc.local and reboot your system At this point your /etc/rc.local should be complete (part 1/3, part 2/3, part 3/3) and the overall system should be ready .
REBOOT and check the process you should observe something like this :
UID PID PPID CMD
...
root 830 1 watchdogd -g dba -e /u01/ora/pro <-- WD
root 831 1 oranm /m /e:/u01/ora/product/8.1 <-- NM
root 832 1 oracm /e:/u01/ora/product/8.1.7. <-- CM
root 844 831 oranm /m /e:/u01/ora/product/8.1
root 845 844 oranm /m /e:/u01/ora/product/8.1
root 851 844 oranm /m /e:/u01/ora/product/8.1 <-- child
root 852 832 oracm /e:/u01/ora/product/8.1.7. threads
root 853 852 oracm /e:/u01/ora/product/8.1.7.
root 854 844 oranm /m /e:/u01/ora/product/8.1
root 855 844 oranm /m /e:/u01/ora/product/8.1
root 856 844 oranm /m /e:/u01/ora/product/8.1
...
6) Parallel database setup
6.1) - Create paralle server database
6.1.1) Prepare the symbolic links
Just to setup a more readable configuration we user OFA template and
symbolic links to Raw Devices . It will look more readable
% ls -l /u01/ora/oradata/OPS/
lrwxrwxrwx 1 oracle dba controlfile -> /dev/raw/raw4
lrwxrwxrwx 1 oracle dba rlog1 -> /dev/raw/raw5
lrwxrwxrwx 1 oracle dba rlog2 -> /dev/raw/raw6
lrwxrwxrwx 1 oracle dba rlog3 -> /dev/raw/raw7
lrwxrwxrwx 1 oracle dba rlog4 -> /dev/raw/raw8
lrwxrwxrwx 1 oracle dba rollback.dbf -> /dev/raw/raw9
lrwxrwxrwx 1 oracle dba system.dbf -> /dev/raw/raw2
lrwxrwxrwx 1 oracle dba temp.dbf -> /dev/raw/raw3
6.1.2) Prepare the init.ora files
Here we will have two instances 'OPS_A' and 'OPS_B' and the respective
private and common parameters files .
Private parameters files :
#--OPS_A private parameters #--OPS_B private parameters # file: ?/dbs/initOPS_A.ora # file: ?/dbs/initOPS_B.ora db_name = "OPS" db_name = "OPS"ifile=$ORACLE_HOME/dbs/initOPS.ora ifile=$ORACLE_HOME/dbs/initOPS.ora
thread=1 thread=2 instance_number=1 instance_number=2 rollback_segments = (RBS1) rollback_segments = (RBS2)
Common parameters file (part of) :
#--OPS Relevant common instance parameters (trivial) # file ?/dbs/initOPS.ora
db_name = "OPS"
control_files = ("/u01/ora/oradata/OPS/controlfile")
compatible ="8.1.7"
parallel_server =true #-- database in Parallel mode
#parallel_server =false #-- database in Standalone mode
gc_rollback_locks ="0-128=32!8REACH" #Default: 0-128=32!8REACH
GC_RELEASABLE_LOCKS =1000
6.1.3) create the database
6.1.4) Redo logs threads
Hint : we suggest to create the initial RLOGS to normal files and then
redefine them to the dedicated raw devices prior to open parallel .
ALTER DATABASE ADD LOGFILE THREAD 1 GROUP 11 '/dev/raw/raw5' size 10M ; ALTER DATABASE ADD LOGFILE THREAD 1 GROUP 12 '/dev/raw/raw6' size 10M ; ALTER DATABASE ADD LOGFILE THREAD 2 GROUP 21 '/dev/raw/raw7' size 10M ; ALTER DATABASE ADD LOGFILE THREAD 2 GROUP 22 '/dev/raw/raw8' size 10M ;ALTER DATABASE ENABLE THREAD 2; (optionally use the 'reuse' clause)
drop the dummy RLOG on normal files
6.1.5) Final checks
Before opening parallel (parallel_server=true) check that :
+) all the WD,NM,CM are running
+) each instance has a private rollback segment created
+) both redo logs thread are active and on raw devices
+) all database is defined on raw devices
7) Startup both the instances in parallel mode
7.1) setup to unix environment in order to have the ORACLE_HOME pointing to the 8.1.7.0.1 code path and two distict ORACLE_SID :
Instance OPS_A Instance OPS_B -------------- -------------- ORACLE_SID=OPS_A ORACLE_SID=OPS_B ORACLE_BASE=/u01/ora ORACLE_BASE=/u01/ora ORACLE_HOME=/u01/ora/product/81701 ORACLE_HOME=/u01/ora/product/81701 ORACLE_TERM=vt100 ORACLE_TERM=vt100 PATH=$PATH:$ORACLE_HOME/bin: PATH=$PATH:$ORACLE_HOME/bin: PATH=$PATH:$ORACLE_HOME/oracm/bin PATH=$PATH:$ORACLE_HOME/oracm/bin LD_LIBRARY_PATH=$ORACLE_HOME/lib:\ LD_LIBRARY_PATH=$ORACLE_HOME/lib:\ $ORACLE_HOME/oracm/lib $ORACLE_HOME/oracm/lib export ORACLE_BASE export ORACLE_BASE export ORACLE_HOME export ORACLE_HOME export ORACLE_TERM export ORACLE_TERM export ORACLE_SID export ORACLE_SID export PATH export PATH export LD_LIBRARY_PATH export LD_LIBRARY_PATH
7.2) startup the instances
The usual symbolic links to init files should already be created in
$ORACLE_HOME/dbs, make sure at this point that 'parallel_server=true'
in common init file.
<OPS_A> <OPS_B> sqlplus / as sysdba sqlplus / as sysdba startup startup ... ...
if all worked ok, both the istances will startup, monitor 'alertOPS_A.log' , 'alertOPS_B.log' for errors
the process view should show the follwing process runnings :
UID PID PPID CMD
...
oracle 1159 1 ora_pmon_OPS_A
oracle 1161 1 ora_lmon_OPS_A
oracle 1163 1 ora_lmd0_OPS_A
oracle 1184 1 ora_dbw0_OPS_A
oracle 1186 1 ora_lgwr_OPS_A
oracle 1188 1 ora_ckpt_OPS_A
oracle 1190 1 ora_smon_OPS_A
oracle 1192 1 ora_reco_OPS_A
oracle 1200 1 ora_lck0_OPS_A
oracle 1203 1 ora_bsp0_OPS_A
...
oracle 1231 1 ora_pmon_OPS_B
oracle 1233 1 ora_lmon_OPS_B
oracle 1235 1 ora_lmd0_OPS_B
oracle 1237 1 ora_dbw0_OPS_B
oracle 1239 1 ora_lgwr_OPS_B
oracle 1241 1 ora_ckpt_OPS_B
oracle 1243 1 ora_smon_OPS_B
oracle 1245 1 ora_reco_OPS_B
oracle 1253 1 ora_lck0_OPS_B
oracle 1256 1 ora_bsp0_OPS_B
...
<end>
8) Conclusions
This document should be intended principally for didactic scopes and used by peoples that are testing the Oracle Parallel Server technology and need to setup a quick solution for simulation, test and developement purposes .
"Marco Mapelli" <mapellim_at_usa.net> wrote in message news:53d76e1f.0110281212.bced917_at_posting.google.com...
> Hello All, > > I am trying with no success to start multiple instances > using a single physical database. > When I mount the second instance I get and error saying that > the database cannot be mounted when in shared mode. > Is there a particular option that let multiple mounting of > an Oracle 8.1.6 DB (RH Linux 6.1). > > Thanks in advance. > > Marco MapelliReceived on Sun Oct 28 2001 - 22:58:59 CST
 |
 |