Oracle® Database Release Notes
10g Release 1 (10.1.0.2.0) for AIX 5L Based Systems (64-Bit) B13611-08 |
|
Release Notes
10g Release 1 (10.1.0.2.0) for AIX 5L Based Systems (64-Bit)
B13611-08
January 2006
This document contains important information that was not included in the platform-specific or product-specific documentation for this release.
To check for updates to this document and to view other Oracle documentation, see the Documentation section on the Oracle Technology Network (OTN) Web site:
http://www.oracle.com/technology/documentation/
For additional information about this release, refer to the readme files located in the $ORACLE_HOME/relnotes
directory.
This document contains the following topics:
The following sections contain information about issues related to Oracle Database 10g and associated products:
Oracle Real Application Clusters Installation and Configuration Guide
Installing Oracle CRS on a System with More Than Three Network Interfaces
Remote Undo Tablespaces Do Not Autoexetend in RAC Seed Databases
Problems Starting Oracle Enterprise Manager Database Control
The Oracle Real Application Clusters Installation and Configuration Guide has been revised. You can download a copy of the revised book, part number B10766-08, from the Documentation section of the OTN Web site:
If you are using raw logical volumes to store Oracle datafiles, Oracle recommends that you use a command similar to the following to create logical volumes with a zero offset:
/usr/sbin/mklv -y LVname -T O -w n -s n -r n VGname NumPPs
Using a zero offset improves database performance and fixes the issues described in Oracle bug 2620053.
Note: On raw logical volumes, if you create tablespaces in datafiles that are not created in this way, a message is recorded in thealert.log file.
|
See Also: For more information about this issue, contact Oracle Support. |
When you install Oracle Database on a system with more than 3 network interfaces, not all of the interfaces are displayed in the menu used to select the public and private interconnects. To work around this problem:
Select public and private interconnects from the three interfaces which are listed in the menu.
Complete the installation.
After the installation, you can change the interconnects by entering the following commands as root
, if necessary:
# oifcfg delif -global # oifcfg setif -global interface/subnet:public # oifcfg setif -global interface/subnet:cluster_interconnect
For example:
# oifcfg delif -global # oifcfg setif -global en0/144.25.68.0:public # oifcfg setif -global en5/192.168.100.0:cluster_interconnect
Enter the following command to verify the new settings:
$ oifcfg getif
When you install Oracle CRS on your system, the private interconnects may not be defined correctly. After installation, enter the following command to verify the public and private interconnects:
$ oifcfg getif
This command should return values for global public
and global cluster_interconnect
, for example:
en0 144.25.68.0 global public en5 192.168.100.0 global cluster_interconnect
If the command does not return a value for global cluster_interconnect
, enter the following commands:
# oifcfg delif -global # oifcfg setif -global interface/subnet:public # oifcfg setif -global interface/subnet:cluster_interconnect
For example:
# oifcfg delif -global # oifcfg setif -global en0/144.25.68.0:public # oifcfg setif -global en5/192.168.100.0:cluster_interconnect
Enter the following command to verify the new settings:
$ oifcfg getif
If you install Oracle CRS on a system that also contains an Oracle9i Real Application Clusters (RAC) cluster, you must restart all of the Oracle CRS nodes. You can restart the nodes one at a time to avoid affecting any Oracle9i databases that may be in use.
Oracle assumes that when you install Oracle CRS 10g software, you will also install Oracle Real Application Clusters 10g. If there will be a significant delay before you perform phase two of the RAC installation to install Oracle Real Application Clusters 10g, then run the /crs_home
/bin/gsdctl start
command to start the GSD manually to service the 9.2 SRVCTL tool and assistants. Then before installing Oracle Real Application Clusters 10g, run the /
crs_home
/bin/gsdctl stop
command to stop the GSD.
If you perform a silent installation of Oracle CRS on multiple nodes, on a system that does not have other Oracle installations, the Installer does not set up the Oracle Inventory correctly. In this case, after the installation is complete, follow these steps:
Run the orainstRoot.sh
script on a local node.
Copy the oraInventory
directory from the local node to each of the remote nodes.
Log in as the root
user and run the following script on each remote node:
oraInventory/orainstRoot.sh
During an Oracle CRS installation, the Installer uses SSH (if available) to run commands and copy files to the other nodes. During the installation, you might see errors similar to the following if a "dot" file on the system (for example, .bashrc
or .cshrc
) contains stty
commands:
stty: standard input: Invalid argument stty: standard input: Invalid argument
To avoid this problem, Oracle recommends that you modify these files to suppress all output on STDERR, as follows:
Bourne, Bash, or Korn shell:
if [ -t 0 ]; then stty intr ^C fi
C shell:
test -t 0 if ($status == 0) then stty intr ^C endif
Note: When SSH is not available, the Installer uses thersh and rcp commands instead of ssh and scp . If there are "dot" files that contain stty commands that are loaded by the remote shell, this error can also occur.
|
This issue is tracked through Oracle bug 3414362.
You must apply Program Technical Fix (PTF) U496549 or higher to your system before you use ASM.
You cannot use the silent installation method to install Oracle Database 10g and create a database that uses Automatic Storage Management (ASM) for database storage during the same installation. This is because the root.sh
script must run before the Database Configuration Assistant (DBCA) can start an ASM instance.
If you want to use the silent installation method to install Oracle Database 10g and create a database that uses ASM, follow these steps:
Use the enterprise.rsp
response file to complete a software-only installation.
Run $ORACLE_HOME/root.sh
after the installation completes.
Use the dbca.rsp
response file to run DBCA in silent mode, using a command similar to the following:
$ $ORACLE_HOME/bin/dbca -silent -responseFile
/full_path/dbca.rsp
To remove the ASM instances, the delete node procedure requires the following additional steps on UNIX-based systems:
If this is the Oracle home from which listener for each node named LISTENER_nodename
runs, then use NetCA to remove this listener and its CRS resources. If necessary, re-create this listener in another Oracle home.
If this is the Oracle home from which the ASM instance runs, then enter the following commands to remove the ASM configuration:
$ srvctl stop asm -n node $ remove asm -n node
If you are using a cluster file system for your ASM Oracle home, then run the following commands on the local node:
$ rm -r $ORACLE_BASE/admin/+ASM $ rm -f $ORACLE_HOME/dbs/*ASM*
If you are not using a cluster file system for your ASM Oracle home, then run the rm
commands listed in the previous step on each node on which the Oracle home exists.
Remove oratab
entries beginning with +ASM
.
After installing Oracle RAC 10g and after ensuring that the system is functioning properly, make a backup of the voting disk. In addition, make a backup of the voting disk contents after you complete any node additions or node deletions and after running any de-installation procedures.
If you create a RAC database with two or more instances and you choose to create General Purpose, Transaction Processing, or Data Warehouse databases, and if you use a shared cluster file system or Automatic Storage Management (ASM) for database files, then the Database Configuration Assistant (DBCA) creates undo tablespace datafiles with an initial size of 25 MB and AUTOEXTEND ON
for the local instance but AUTOEXTEND OFF
for remote instances.
You can set AUTOEXTEND ON
for undo tablespace datafiles for remote instances after creating a RAC database using DBCA as follows:
Connect to the database instance on the node from which you ran DBCA:
$ sqlplus "/ AS SYSDBA"
Enter the following command to find the datafile names for UNDOTBS
tablespaces for remote instances:
SQL> SELECT file_name FROM SYS.DBA_DATA_FILES WHERE tablespace_name LIKE 'UNDOTBS%' AND AUTOEXTENSIBLE='NO';
Set AUTOEXTEND ON
for the datafiles that you found in the previous step:
SQL> ALTER DATABASE DATAFILE datafile_name AUTOEXTEND ON;
If you choose the Custom installation path and have previous versions of Oracle databases installed, then the Oracle Universal Installer gives you the option of upgrading the existing databases. If you choose to upgrade an existing database, the Oracle Universal Installer displays another screen asking whether you want to create a starter database.If you choose to create a starter database, the DBCA upgrades the older version of the database to Oracle Database 10g release 1 (10.1), but does not create a starter database.
If you upgrade an Oracle9i database to Oracle Database 10g release 1, Oracle Flashback features using a timestamp may fail. To work around this problem, enter the following SQL script from the Oracle Database 10g database:
SQL> DELETE FROM smon_scn_time WHERE orig_thread <> 0; SQL> COMMIT;
This issue is tracked with Oracle bug 3994270.
In the Net Configuration Assistant (NetCA) help, the link to the Select Oracle Context help topic is broken. The text for this topic is as follows:
Directory Usage Configuration, Select Oracle Context
Oracle administrative content has been found in more than one location in the directory. Oracle administrative content is stored in an Oracle Context, a subtree in the directory that stores Oracle entries.
From the list, select or enter the location you want to use as the default Oracle Context location from which this computer will access Oracle entries, such as connect identifiers.
If a user invokes the Flashback Table or Flashback Analysis operation, and that user has FLASHBACK ANY TABLE privileges but does not have specific flashback privileges on the objects that flashback is invoked on and does not have DBA privileges, then the following errors may occur:
ORA-02002: error while writing to audit trail ORA-00600: internal error code, arguments: [kzasps1], [4], [47], [],[],
To fix this problem, as SYSDBA grant the user FLASHBACK privilege on the objects that are referred to in the FLASHBACK TABLE statement and then invoke the flashback operation.
For example:
SQL> GRANT FLASHBACK ON SCOTT.EMP_1 TO user1;
This issue is tracked through Oracle bug 3403666.
Although Oracle Workflow is listed on the Companion CD installation screens, it is not included on the Companion CD for this release.
While installing Oracle Database, the Specify Backup and Recovery Options screen may appear truncated if your system does not have the required fonts installed. If your system has only fixed-width fonts, you may not be able to fully specify the required information in the Backup Job Credentials area of the screen. To work around this issue, do not select Enable Automated Backups on this screen. After the installation is complete, use the Oracle Enterprise Manager 10g Database Control to enable automated backups.
This release includes the Oracle Internet Directory (OID) client tools, but it does not include OID server components. OID server components are included with Oracle Application Server 10g. If you require the OID server tools for Oracle Database components, then run them from an Oracle Application Server 10g installation.
The OID client tools include:
LDAP command-line tools
Oracle Internet Directory SDK
Oracle Directory Manager
The OID server components include the following servers and tools for starting and stopping them:
Directory server
Directory replication server
Directory integration server
Viewing the execution plan of a Period SQL in Korean causes an internal server error. This problem is unique to Korean; it does not reproduce in Japanese or Chinese. The only workaround currently available is to run the product in a language other than Korean when you need to view this page.
Quick Tour is not available in Oracle Change Management Pack. If you try to run it, then an error results.
To enable the extjob
executable to locate required libraries, the $ORACLE_HOME/lib
directory and all of its parent directories must have execute permissions for group
and other
.
Threads in AIX can run with process-wide contention scope (M:N) or with systemwide contention scope (1:1). The AIXTHREAD_SCOPE environment variable controls which contention scope is used.
The default value of the AIXTHREAD_SCOPE environment variable is P which specifies process-wide contention scope. When using process-wide contention scope, Oracle threads are mapped to a pool of kernel threads. When Oracle is waiting on an event and its thread is swapped out, it may return on a different kernel thread with a different thread ID. Oracle uses the thread ID to post waiting processes so it is important for the thread ID to remain the same. When using systemwide contention scope, Oracle threads are mapped to kernel threads statically, one to one. For this reason Oracle recommends using systemwide contention. The use of systemwide contention is especially critical for Oracle Real Application Clusters (RAC) instances.Additionally, on AIX 5L version 5.2 or higher, if you set systemwide contention scope, significantly less memory is allocated to each Oracle process.
Oracle recommends that you set the value of the AIXTHREAD_SCOPE environment variable to S in the environment script that you use to set the ORACLE_HOME or ORACLE_SID environment variables for an Oracle database instance or an Oracle Net listener process, as follows:
Bourne, Bash, or Korn shell:
Add the following line to the ~/.profile
or /usr/local/bin/oraenv
script:
AIXTHREAD_SCOPE=S; export AIXTHREAD_SCOPE
C shell:
Add the following line to the ~/.login
or /usr/local/bin/coraenv
script:
setenv AIXTHREAD_SCOPE S
Doing this enables systemwide thread scope for running all Oracle processes.
Authorized Problem Analysis Report (APAR) IY56024 is required to prevent concurrent I/O writes returning an incorrect length on AIX 5L systems.
See Also: For more information on this patch, see the following Web site:http://www-912.ibm.com/eserver/support/fixes/ |
Oracle Transparent Gateways for Sybase does not currently work with Sybase Open Client V12.0. A patch to correct this problem is available from the Download area of the OracleMetalink Web site:
http://metalink.oracle.com/
This issue is tracked through Oracle bug 3398064.
If you choose to configure Oracle Enterprise Manager Database Control during an Oracle Database installation on AIX 5L version 5.3, Database Control might not start. This problem occurs because the Installer cannot obtain a free port in the default range 1830 to 1849.
To work around this problem, before installation edit the /etc/services
file to remove or comment out the port assignments for ports in the range 1830 to 1849.
If you do not edit the /etc/services
file before installation, perform the following steps after installation:
Open the following file in a text editor:
$ORACLE_HOME/hostname_sid/sysman/config/emd.properties
Add an available TCP port to the value of the EMD_URL parameter as follows:
EMD_URL=http://hostname:available_tcp_port/emd/main
Open the following file in a text editor:
$ORACLE_HOME/hostname_sid/sysman/emd/targets.xml
Edit the port number in the oracle_emd
target type to match the port number that you specified in step 2.
Restart Database Control as follows:
$ emctl start dbconsole
Before installing Oracle Database or before starting Oracle Enterprise Manager Database Control, you must set the DISPLAY environment variable to specify a valid X server as follows, where workstation_name
is the name of a valid X server:
Bourne, Bash, or Korn shell:
$ DISPLAY=workstation_name:0.0 ; export DISPLAY
C shell:
% setenv DISPLAY workstation_name:0.0
If the DISPLAY environment variable specifies a server that is not an X server, the Installer fails or Oracle Enterprise Manager Database Control produces Java core files and users will not be able to connect to it from the browser.
Oracle Database New Features for Oracle Database 10g release 1 (10.1) lists two Grid features that are not available in the first release of Oracle Database 10g; Resonance and Transparent Session Migration. These features will be available in a future release.
Note: This issue occurs only if the cloned and source Oracle homes are on the same system. |
If you use the Installer to remove a cloned Oracle home, and the cloned Oracle home is on the same system as the source Oracle home, the Installer removes essential files from the source Oracle home. To avoid this issue, remove the cloned Oracle home as follows:
Remove the cloned Oracle home manually by deleting the cloned Oracle home directory.
Start the Installer from the source Oracle home directory:
$ source_oracle_home/oui/bin/runInstaller
In the Welcome window, click Deinstall Products.
The Inventory window appears, listing all of the Oracle homes on the system.
In the Inventory window, select the cloned Oracle home, then click Remove.
This issue is tracked through Oracle bug 3417663.
If you use the JDBC driver, and the JAVA_COMPILER environment variable is set to any value other than NONE, Jserver will not start. This problem is fixed by setting JAVA_COMPILER to NONE
. For example:
Set the JAVA_COMPILER environment variable.
Bourne, Bash, or Korn shell:
$ JAVA_COMPILER=NONE ; export JAVA_COMPILER
C shell:
% setenv JAVA_COMPILER NONE
Specify the following option with the java
command:
-Djava.compiler=none
You cannot use Oracle Scheduler to run jobs external to the database on your system.
To install Enterprise Security Manager (ESM), install Oracle Client and choose the Administrator installation type.
When using IBM JDK 1.4.2 to build an application that uses the org.w3c.dom.ls
package, the following errors might be thrown:
cannot resolve symbol symbol : method getConfig () location: interface org.w3c.dom.ls.LSParser
or
cannot resolve symbol symbol : method getConfig () location: interface org.w3c.dom.ls.LSSerializer
This issue is being tracked with Oracle bug 4286338.
The following sections contain updates to Oracle documentation:
The following sections describe corrections and supplemental information for the Oracle Real Application Clusters 10g documentation for UNIX platforms:
The Oracle Real Application Clusters Installation and Configuration Guide, Chapter 8, "Installing Cluster Ready Services", under the heading "Installing Cluster Ready Services with the OUI", explains how to run the following script before running Oracle Universal Installer from the Oracle CRS CD-ROM:
ORACLE_HOME/bin/localconfig delete
Instead, run this script after starting the Oracle CRS installation when Oracle displays the warning to stop all Oracle services.
Note: The documentation is correct by instructing you to stop the existing ASM instances before installing Oracle CRS. Only the timing of when to runlocalconfig delete is changed from before to during the installation.
|
If you are installing RAC on an AIX cluster without HACMP or GPFS, you must use shared raw disk devices to store the Cluster Ready Services (CRS) files. This section describes how to configure the shared raw disk devices for the CRS files (Oracle Cluster Registry and CRS voting disk) and the database files.
Note: You can also use shared raw disk devices to store database files. However, Oracle recommends that you use ASM for database file storage if you are not using HACMP or GPFS.To configure shared raw disk devices for database files, configure the required shared devices as described in steps 2 through 7. The number and size of the required raw devices are listed in Step 2 of the "Create Raw Logical Volumes in the New Volume Group" section in the Oracle Real Application Clusters Installation and Configuration Guide. |
The Oracle Cluster Registry (OCR) requires a disk with 100 MB of storage space and the CRS voting disk requires a disk with 20 MB of storage space.
Note: Because each file requires exclusive use of a complete disk device, Oracle recommends that you use this storage option for CRS files only if you do not have HACMP or GPFS. You cannot use the disks that you choose for these files for any other purpose. |
To configure raw shared disk devices for the OCR and CRS voting disk, follow these steps:
Select two unused disks that are shared on all of the cluster nodes.
As the root
user, enter the following command on any node to identify the device names for the physical disks that you want to use:
# /usr/sbin/lspv | grep -i none
This command displays information similar to the following for each disk that is not configured in a volume group:
hdisk17 0009005fb9c23648 None
In this example, hdisk17
is the device name of the disk and 0009005fb9c23648
is the physical volume ID (PVID).
If the disks that you want to use for the CRS files do not have a PVID, enter a command similar to the following for each disk that you want to use:
# /usr/sbin/chdev -l hdiskn -a pv=yes
On each of the other nodes, enter a command similar to the following to identify the device names associated with the PVID on that node:
# /usr/sbin/lspv | grep -i "0009005fb9c23648"
The output from this command should be similar to the following:
hdisk18 0009005fb9c23648 None
In this example, the device name associated with the drive (hdisk18
) is different on this node.
If the device names are the same on all nodes, enter commands similar to the following to change the owner, group, and permissions on the character raw device files for the disks:
# chown oracle:dba /dev/rhdiskn # chmod 660 /dev/rhdiskn
If the device name associated with the PVID for a disk that you want to use is different on any node, you must create a new device file for the disk on each of the nodes using a common unused name.
For the new device file, choose a device file name that identifies the purpose of the disk device, for example ocr_disk
or crs_vote_disk
.
Note: Alternatively, you could choose a name that contains a number that will never be used on any of the nodes, for examplehdisk99 .
|
To create a new common device file name for a disk on all nodes, follow these steps on each node:
Enter the following command to determine the device major and minor numbers that identify the disk, where n
is the disk number for the disk on this node:
# ls -alF /dev/*hdiskn
The output from this command is similar to the following:
brw------- 1 root system 24,8192 Dec 05 2001 /dev/hdiskn crw------- 1 root system 24,8192 Dec 05 2001 /dev/rhdiskn
In this example, the device file /dev/rhdisk
n
represents the character raw device, 24
is the device major number, and 8192
is the device minor number.
Enter a command similar to the following to create the new device file, specifying the new device file name and the device major and minor numbers that you identified in the previous step:
Note: In the following example, you must specify the characterc to create a character raw device file.
|
# mknod /dev/ocr_disk c 24 8192
Enter commands similar to the following to change the owner, group, and permissions on the character raw device file for the disk:
# chown oracle:dba /dev/ocr_disk # chmod 660 /dev/ocr_disk
Enter a command similar to the following to verify that you have created the new device file successfully:
# ls -alF /dev |grep "24,8192"
The output should be similar to the following:
brw------- 1 root system 24,8192 Dec 05 2001 /dev/hdiskn crw-rw---- 1 oracle dba 24,8192 Dec 05 2001 /dev/ocr_disk crw------- 1 root system 24,8192 Dec 05 2001 /dev/rhdiskn
To enable simultaneous access to a disk device from multiple nodes, you must set the appropriate Object Data Manager (ODM) attribute listed in the following table to the value shown, depending on the disk type:
Disk Type | Attribute | Value |
---|---|---|
EMC PowerPath devices, SSA, or FAStT disks | reserve_lock | no
|
ESS, HDS, or MPIO-capable disks | reserve_policy | no_reserve
|
To determine whether the attribute has the correct value, do one of the following depending on your disk type:
SSA, FAStT, ESS, HDS, or MPIO-capable disks:
Enter a command similar to the following on all cluster nodes for each disk device that you want to use:
# /usr/sbin/lsattr -E -l hdiskn
EMC PowerPath devices:
Enter a command similar to the following on each of the CRS and database devices that you want to use:
# /usr/sbin/lsattr -E -l hdiskpowern
If necessary, change the attribute value as follows, depending on the disk type:
SSA or FAStT disks:
# /usr/sbin/chdev -l hdiskn -a reserve_lock=no
EMC PowerPath devices:
# /usr/sbin/chdev -l hdiskpowern -a reserve_lock=no
ESS, HDS, or MPIO-capable disks:
# /usr/sbin/chdev -l hdiskn -a reserve_policy=no_reserve
Note: If the appropriate attribute is not set to the value listed in the table, problems may occur during access from multiple nodes. |
If necessary, enter the following command to remove all physical volume IDs (PVIDs) from the raw devices:
# /usr/sbin/chdev -l hdiskn -a pv=clear
When you are installing Oracle CRS, enter the paths to the appropriate device files when prompted for the path of the OCR and CRS voting disks, for example:
/dev/ocr_disk
The following points describe corrections for the node addition procedures in Chapter 5, "Adding and Deleting Nodes and Instances" in the Oracle Database Oracle Clusterware and Oracle Real Application Clusters Administration and Deployment Guide.
Under the heading "Step 2 Adding Nodes at the Oracle Clusterware Layer (UNIX and Windows)", point 9 on page 5-6, the format for the racgons
command should be:
racgons add_config new_node_name:4948
Under the heading "Step 4: Adding Nodes at the Oracle RAC Database Layer", point 8 on page 5-9, explains that you must run the Virtual Internet Protocol Assistant (VIPCA). You must run the VIPCA with root
user privileges.
Under the heading "Step 4: Adding Nodes at the Oracle RAC Database Layer", point 9 on page 5-10 explains that you must run the oifcfg
command. The correct syntax for this command requires that you specify either the -n nodename
option or the -global
option to complete the command. The correct syntax for the oifcfg
command is as follows:
oifcfg iflist oifcfg setif {-node nodename | -global} {if_name/subnet:if_type}... oifcfg getif [-node nodename | -global] [ -if if_name[/subnet] [-type if_type] ] oifcfg delif [-node nodename | -global] [if_name[/subnet]] oifcfg [-help]
A standard Oracle CRS installation issues the oifcfg
command as in the following example:
oifcfg setif -global eth0/146.56.76.0:public \ eth1/192.0.0.0:cluster_interconnect'
This sets both networks to global
. Therefore, you do not need to run the oifcfg
command manually after you add a node.
When you add a node to a RAC database, you must also add a listener to the new node by running the Net Configuration Assistant (NetCA).
The following corrections are for the node deletion procedures for UNIX-based systems as described in Chapter 5, "Adding and Deleting Nodes and Instances" in the Oracle Database Oracle Clusterware and Oracle Real Application Clusters Administration and Deployment Guide. These corrections assume that you have a three-node cluster and that you are deleting the third node.
Step 2 in the node deletion procedure for UNIX-based systems should read:
After you have deleted the instances from the nodes that you want to delete, delete the node applications for each node by running the following command where <node1, node2>
are the nodes that you are removing from your cluster:
rootdeletenode.sh <node1>,<node2>
For Step 3 in the node deletion procedure for UNIX-based systems, the runInstaller
command is located in the $ORACLE_HOME/oui/bin directory
. In addition, executing this command does not launch the Installer.
For Step 5 in the node deletion procedure for UNIX-based systems, only run the <CRS Home>/install/rootdelete.sh
command once for that step. Use the argument nosharedhome
when using local file systems. The default is sharedhome
which prevents updating the permissions of local files such that they can be removed by the oracle
user.
For Step 7 in the node deletion procedures for UNIX-based systems, the command is located in the Oracle CRS Home directory. In this step, ORACLE_HOME
should read CRS_Home
and the text for Step 7 should read:
On the same node and as the oracle
user, run the command <CRS home>/oui/bin/runInstaller -updateNodeList ORACLE_HOME=<CRS home> CLUSTER_NODES=node1,node2,... <nodeN>
where node1
through <nodeN>
is a comma-delimited list of nodes that are remaining in the cluster.
The following two sections that describe shadow events and high availability callouts with Oracle Notification Services (ONS) events are supplements for Appendix A, "Services Deployment Example", in the Oracle Real Application Clusters Deployment and Performance Guide, part number B10768-01.
When using Transparent Application Failover (TAF) PRECONNECT, Real Application Clusters (RAC) high availability maintains a preconnect service to support TAF Preconnect and applications that are configured to manage work on secondary RAC instances. Secondary instances are RAC instances that are not supporting the primary service.
In this type of configuration, Oracle maintains the shadow service on all instances that do not support the primary service. You can use events to stop and start secondary work. The events are posted to callouts and to the Oracle Notification Service (ONS).
To use events, configure the payload with the following format:
Up Event Format:
SRV_PRECONNECT VERSION=1.0 service=db_unique_name.db_domain database=database_name instance instance_name host=host_name status=preconn_up reason=timestamp=27-Jan-2004 16:53:58 reported=Tue Jan 27 16:53:59 PST 2004
Down Event Format:
SRV_PRECONNECT VERSION=1.0 service=db_unique_name.db_domain database=database_name instance=instance_name host=host_name status=preconn_down reason=timestamp=27-Jan-2004 16:58:01 reported=Tue Jan 27 16:58:02 PST 2004
Up Event Example:
@ SRV_PRECONNECT VERSION=1.0 service=MYSERV.us.oracle.com database=RACEY instance=RACEY1 host=sun880-1 status=preconn_up reason=timestamp=27-Jan-2004 16:53:58 reported=Tue Jan 27 16:53:59 PST 2004
Down Event Example:
@ SRV_PRECONNECT VERSION=1.0 service=MYSERV.us.oracle.com database=RACEY instance=RACEY1 host=sun880-1 status=preconn_down reason=timestamp=27-Jan-2004 16:58:01 reported=Tue Jan 27 16:58:02 PST 2004
The notification interface is available as a server-side callout and as an Oracle Notification Services (ONS) event. The server-side callout is a script with the same payload as the ONS event that is immediately executed on the server when the condition occurs. Use this method to start and stop server-side applications, to relocate low-priority services when high priority services arrive, and to post tickets for fault tracking. The following table describes the event payload:
Parameter | Description |
---|---|
Event type | The event type for the component such as service, service_member, database, instance, or node |
Service name | The service name; matches the configured service in SERVICE$ |
Database name | The database supporting the service; matches the initialization parameter value for DB_UNIQUE_NAME, which in turn defaults to the value of the initialization parameter DB_NAME |
Instance | The name of the instance that supports the service; matches the instance name |
Node name | The name of the node that supports the service or the node that has failed; matches the CSS node name |
Status | The new status; values are UP, DOWN, and NOT_RESTARTING |
Cardinality | Cardinality for the service on UP events |
Time stamp | The local time zone to use when ordering notification events |
Incarnation | Cluster incarnation for node down |
When a session connects, mid-tiers can record the following values that match the HA event payload:
sys_context('userenv', 'instance_name'); sys_context('userenv', 'server_host'); sys_context('userenv', 'service_name'); sys_context('userenv', 'db_unique_name');
If you intend to use ASM as the storage mechanism for a RAC database, configure the raw disk devices used in the ASM disk group to enable simultaneous access from multiple nodes. To enable simultaneous access to a disk device from multiple nodes, you must set the appropriate Object Data Manager (ODM) attribute listed in the following table to the value shown, depending on the disk type:
Disk Type | Attribute | Value |
---|---|---|
EMC PowerPath devices, SSA, or FAStT disks | reserve_lock | no
|
ESS, HDS, or MPIO-capable disks | reserve_policy | no_reserve
|
To determine whether the attribute has the correct value, do one of the following depending on your disk type:
SSA, FAStT, ESS, HDS, or MPIO-capable disks:
Enter a command similar to the following on all cluster nodes for each disk device that you want to use:
# /usr/sbin/lsattr -E -l hdiskn
EMC PowerPath devices:
Enter a command similar to the following on each of the CRS and database devices that you want to use:
# /usr/sbin/lsattr -E -l hdiskpowern
If the required attribute is not set to the correct value on any node, enter a command similar to one of the following on that node:
SSA or FAStT disks
# /usr/sbin/chdev -l hdiskn -a reserve_lock=no
EMC PowerPath devices
# /usr/sbin/chdev -l hdiskpowern -a reserve_lock=no
ESS, HDS, or MPIO-capable disks
# /usr/sbin/chdev -l hdiskn -a reserve_policy=no_reserve
Before raw devices are used, remove all physical volume IDs (PVIDs) from the raw devices. PVIDs may have been assigned by the user to correlate the hdisk numbers for disks shared across RAC nodes. Enter the following command to remove the PVID.
# /usr/sbin/chdev -l hdiskn -a pv=clear
Enter commands similar to the following on every node to change the owner, group, and permissions on the character raw device file for each disk that you want to add to the disk group:
# chown oracle:dba /dev/rhdiskn # chmod 660 /dev/rhdiskn
Note: If you are using a multi-pathing disk driver with ASM, make sure that you set the permissions only on the correct logical device name for the disk. The device name associated with a disk might be different on other nodes. Make sure that you specify the correct device name on each node. |
The following text appears on page 5-10 of the Oracle Real Application Clusters Administrator's Guide:
Execute the following procedures on each new node to add instances:
Start the Database Configuration Assistant (DBCA) by entering dbca
at the system prompt from the bin
directory in the $ORACLE_HOME
on UNIX.
This text should be changed as follows:
Execute the following procedures for each new node to add instances:
In Oracle Real Application Clusters Installation and Configuration Guide10g Release 1 (10.1) for AIX-Based Systems, Apple Mac OS X, hp HP-UX, hp Tru64 UNIX, Linux, Solaris Operating System, and Windows Platforms, srvclt
is incorrectly spelled as svrctl
in the following sections:
Chapter 11, "Installing Oracle Database 10g with Real Application Clusters"
Chapter 13, "Real Application Clusters Post-Installation Procedures"
Appendix B, "Using Scripts to Create Real Application"
The following sections contain updates to the Oracle Database 10g documentation:
When entering the commands shown in the "Pro*C/C++ Precompiler" section of the Oracle Database 10g Administrator's Reference, include the -r
option, for example:
$ make -r -f demo_proc.mk sample1
If you do not use the -r
option, the demonstrations may not compile correctly.
Although the Oracle Database Companion CD Installation Guide states that you can install Oracle HTML DB into an existing Oracle home that contains Oracle Application Server 10g or Oracle Enterprise Manager 10g, this type of installation is not supported with this release. You must install Oracle HTML DB with the Apache Standalone component in a separate Oracle home directory.
The instructions about the ORA_NLS10 environment variable in the "Using Oracle9i Database Language and Territory Definition Files with Oracle Database 10g" section of the Oracle Database 10g README file are incorrect. This is because the $ORACLE_HOME/nls/data/old
directory contains definitions files which have changed in Oracle Database 10g release 1 only, and not the complete set of the database locale definition files. To include the Oracle9i definitions:
Create a new directory:
$ mkdir $ORACLE_HOME/nls/data/9i
Copy the contents of the $ORACLE_HOME/nls/data
directory to the new directory:
$ cp $ORACLE_HOME/nls/data/*.nlb $ORACLE_HOME/nls/data/9i
Copy the contents of the $ORACLE_HOME/nls/data/old
directory to the directory that you created in step 1:
$ cp $ORACLE_HOME/nls/data/old/*.* $ORACLE_HOME/nls/data/9i
Set the ORA_NLS10 environment variable to specify the directory that you created, for example:
Bourne, Bash, or Korn shell:
$ ORA_NLS10=$ORACLE_HOME/nls/data/9i ; export ORA_NLS10
C shell:
% setenv ORA_NLS10 $ORACLE_HOME/nls/data/9i
See Also: Refer to the Oracle Database 10g Globalization Support Guide and the Oracle Database 10g README for more information about the changes in the language and territory definition files between Oracle9i and Oracle Database 10g. |
For full-text searching with Oracle Text, you must create XML tables manually.
If you will need to use Oracle Text indexes for text-based ora:contains
searches over a collection of XML elements, then do not use XML schema annotation storeVarrayAsTable="true"
. This annotation causes element collections to be persisted as rows in an Index Organized Table (IOT). Oracle Text does not support IOTs.
To be able to use Oracle Text to search the contents of element collections, set parameter genTables="false"
during schema registration. Then create the necessary tables manually, without using the clause ORGANIZATION INDEX OVERFLOW
. The tables will then be heap-organized instead of index-organized (IOT), as shown in the following example:
CREATE TABLE PurchaseOrder of XMLTYPE XMLSCHEMA http://localhost:8080/home/SCOTT/poSource/xsd/purchaseOrder.xsd ELEMENT "PurchaseOrder" VARRAY "XMLDATA"."ACTIONS"."ACTION" STORE AS TABLE ACTION_TABLE ((PRIMARY KEY (NESTED_TABLE_ID, ARRAY_INDEX))) VARRAY "XMLDATA"."LINEITEMS"."LINEITEM" STORE AS TABLE LINEITEM_TABLE ((PRIMARY KEY (NESTED_TABLE_ID, ARRAY_INDEX)));
Oracle Database Performance Tuning Guide, Chapter 10, "Instance Tuning Using Performance Views", has two incorrect view descriptions. The view names and correct descriptions are as follows:
V$EVENT_HISTOGRAM
The V$EVENT_HISTOGRAM view displays a histogram of the number of waits, the maximum wait, and total wait time on an event basis.
V$SYSTEM_WAIT_CLASS
The V$SYSTEM_WAIT_CLASS view provides the instance-wide time totals for the number of waits and the time spent in each class of wait events. (This description removes the reference to object number.)
In Oracle High Availability Architecture and Best Practices, Chapter 8, "Using Enterprise Manager for Monitoring and Detection", the section "Configure At Least Two Service Processes and Load Balance Them" should read as follows: For the middle tier, the baseline recommendation is to have a minimum of two Management Service processes, using a hardware server load balancer to mask the location of an individual Management Service process and a failure of any individual component. This provides immediate coverage for a single failure in the most critical components in the Enterprise Manager architecture with little interruption of service for all systems monitored using Enterprise Manager. Management Service processes connect to the repository instances using Oracle Net.
In Oracle Database Globalization Support Guide, Chapter 5, "Linguistic Sorting and String Searching", the section called "Monolingual Linguistic Sorts" should end with the following note:
Monolingual linguistic sorting is not available for nonUnicode multibyte database character sets. If a monolingual linguistic sort is specified when the database character set is nonUnicode multibyte, then the default sort order is the binary sort order of the database character set. One exception is UNICODE_BINARY. This sort is available for all character sets.
XDK error messages are available at the XML Technology Center on the OTN Web site:
http://www.oracle.com/technology/tech/xml/doc/production10g/Javaerrormsgs.html
Our goal is to make Oracle products, services, and supporting documentation accessible, with good usability, to the disabled community. To that end, our documentation includes features that make information available to users of assistive technology. This documentation is available in HTML format, and contains markup to facilitate access by the disabled community. Accessibility standards will continue to evolve over time, and Oracle is actively engaged with other market-leading technology vendors to address technical obstacles so that our documentation can be accessible to all of our customers. For more information, visit the Oracle Accessibility Program Web site at http://www.oracle.com/accessibility/
.
Accessibility of Code Examples in Documentation
Screen readers may not always correctly read the code examples in this document. The conventions for writing code require that closing braces should appear on an otherwise empty line; however, some screen readers may not always read a line of text that consists solely of a bracket or brace.
Accessibility of Links to External Web Sites in Documentation
This documentation may contain links to Web sites of other companies or organizations that Oracle does not own or control. Oracle neither evaluates nor makes any representations regarding the accessibility of these Web sites.
TTY Access to Oracle Support Services
Oracle provides dedicated Text Telephone (TTY) access to Oracle Support Services within the United States of America 24 hours a day, seven days a week. For TTY support, call 800.446.2398.
Oracle Database Release Notes, 10g Release 1 (10.1.0.2.0) for AIX 5L Based Systems (64-Bit)
B13611-08
Copyright © 1996, 2006, Oracle. All rights reserved.
The Programs (which include both the software and documentation) contain proprietary information; they are provided under a license agreement containing restrictions on use and disclosure and are also protected by copyright, patent, and other intellectual and industrial property laws. Reverse engineering, disassembly, or decompilation of the Programs, except to the extent required to obtain interoperability with other independently created software or as specified by law, is prohibited.
The information contained in this document is subject to change without notice. If you find any problems in the documentation, please report them to us in writing. This document is not warranted to be error-free. Except as may be expressly permitted in your license agreement for these Programs, no part of these Programs may be reproduced or transmitted in any form or by any means, electronic or mechanical, for any purpose.
If the Programs are delivered to the United States Government or anyone licensing or using the Programs on behalf of the United States Government, the following notice is applicable:
U.S. GOVERNMENT RIGHTS Programs, software, databases, and related documentation and technical data delivered to U.S. Government customers are "commercial computer software" or "commercial technical data" pursuant to the applicable Federal Acquisition Regulation and agency-specific supplemental regulations. As such, use, duplication, disclosure, modification, and adaptation of the Programs, including documentation and technical data, shall be subject to the licensing restrictions set forth in the applicable Oracle license agreement, and, to the extent applicable, the additional rights set forth in FAR 52.227-19, Commercial Computer Software—Restricted Rights (June 1987). Oracle Corporation, 500 Oracle Parkway, Redwood City, CA 94065
The Programs are not intended for use in any nuclear, aviation, mass transit, medical, or other inherently dangerous applications. It shall be the licensee's responsibility to take all appropriate fail-safe, backup, redundancy and other measures to ensure the safe use of such applications if the Programs are used for such purposes, and we disclaim liability for any damages caused by such use of the Programs.
Oracle, JD Edwards, PeopleSoft, and Retek are registered trademarks of Oracle Corporation and/or its affiliates. Other names may be trademarks of their respective owners.
The Programs may provide links to Web sites and access to content, products, and services from third parties. Oracle is not responsible for the availability of, or any content provided on, third-party Web sites. You bear all risks associated with the use of such content. If you choose to purchase any products or services from a third party, the relationship is directly between you and the third party. Oracle is not responsible for: (a) the quality of third-party products or services; or (b) fulfilling any of the terms of the agreement with the third party, including delivery of products or services and warranty obligations related to purchased products or services. Oracle is not responsible for any loss or damage of any sort that you may incur from dealing with any third party.