Oracle® Beehive Installation Guide Release 1 (1.5) for Solaris Operating System (SPARC 64-Bit) Part Number E14832-05 |
|
|
View PDF |
This section describes Oracle Beehive database requirements, creating databases, and configuring database settings. It covers the following topics:
Note:
The installation scenario "Installing Oracle Beehive for Demilitarized Zone (DMZ)" described in "What You Should Know Before Installing Oracle Beehive" does not require a database. Refer to this section for more information about this scenario.Oracle Beehive requires an existing database with the following characteristics:
Uses one of the follwing databases installed on any supported operating system:
Oracle Database 10g Release 2 (10.2.0.1) Enterprise Edition with Oracle Database 10g Release 2 (10.2.0.4) Patch Set 3
Oracle Database 11g Release 1 (11.1.0.6) Enterprise Edition
Oracle Database 11g Release 1 (11.1.0.7) Enterprise Edition
Ensure that you have applied all the patches listed in the section "Patches".
Uses character set AL32UTF8 (Unicode)
Contains the following standard database components:
Oracle XML DB
Oracle interMedia (Oracle Multimedia in Oracle Database 11g)
Oracle Text
Partitioning
Contains Rules Manager and Expression Filter (Rules Manager and Expression Filter are installed automatically with Oracle Database 10g Enterprise Edition.)
Microsoft Windows
For Microsoft Windows (32-bit and 64-bit versions), install the latest patch bundle for Oracle Database 10g Release 2 (10.2.0.4), Oracle Database 11g Release 1 (11.1.0.6), or Oracle Database 11g Release 1 (11.1.0.7) Enterprise Edition.
Operating Systems Other Than Microsoft Windows
Install the following patches for Oracle Database 10g Release 2 (10.2.0.4):
6369463: WRONG RESULTS WHEN USING ORA_ROWSCN PSEUDOCOLUMN
4033868: COLLECT FUNCTION LEAVES TEMPORARY SYS TYPES BEHIND
6653934: XF11.1SSHR - TRC - KDRWRIC
6725634: STBH SYNONYMS ARE INVALIDATED WHEN A TABLE PARTITION IS DROPPED
6897966: JOBS AREN'T STARTED SERVICE NAME IS CASE SENSITIVE
6923450: ORA-07445 [KOTGTSCH] - AFTER UPGRADE TO 10.2.0.4
6145687: BETA4:VERY STRANGE SYSTEM CHANGE NUMBER
6782437: QUERY WITH MULTIPLE SUBQUERIES PERFORMING POORLY AFTER 10G UPGRADE
6083201: MERGE STATEMENT WITH XML TYPE FAILS WITH ORA-7445 [FCLOSE()+29]
6144426: STARSQL LNX ORA-04030 WHEN TRYING TO ALLOCATE 1332 BYTES (KXS-HEAP-C,QKKELE)
7694979: ORA-00904: RAISED FOR COMPLEX ADTS RESIDING IN DIFF. SCHEMAS
8214576: JOBS AREN'T STARTED: SERVICE NAME IS CASE SENSITIVE
Install the following patches for Oracle Database 11g Release 1 (11.1.0.6):
6168363: UTL_RECOMP.RECOMP_SERIAL(..) DUMPS WITH ORA 7445 [ PHF_COPY_TO_DIANA()+1329 ]
6708565: EXFSYS.DBMS_RLMGR FAILS WITH ORA-38500
6750049: LIBNNZ10.SO DLOPEN FAILED WHEN TRYING TO START ONS (NODEAPPS)
6526468: VIRTUAL BYTES OF ORACLE.EXE GROWS WHEN INSERT/DELETE REPEATED WITH ONE SESSION
Install the following patches for Oracle Database 11g Release 1 (11.1.0.7):
6782437: QUERY WITH MULTIPLE SUBQUERIES PERFORMING POORLY AFTER 10G UPGRADE
7273988: TST&PERF:QUERY PLAN IS VERY BAD IN RDBMS_MAIN_LINUX_080623
6750049: LIBNNZ10.SO DLOPEN FAILED WHEN TRYING TO START ONS (NODEAPPS)
6083201: MERGE STATEMENT WITH XML TYPE FAILS WITH ORA-7445 [FCLOSE()+29]
6977167: ORA-04092: CANNOT ROLLBACK IN A TRIGGER RECEIVED REQUIRING INSTANCE BOUNCE
7694979: ORA-00904: RAISED FOR COMPLEX ADTS RESIDING IN DIFF. SCHEMAS
7378322: ORA 600 - 6704 GENERATED FOR A JOIN QUERY ON A HASH PARTITIONED TABLE
8221425: ORA-04031: UNABLE TO ALLOCATE 912 BYTES OF SHARED MEMORY
7643188: SQL VERSION COUNT REPORTED IN AWR DOES NOT MATCH SQLSTATS FOR 1 SQL ID
7156912: ORACLE TEXT INDEXING HANGS ON A HTML DOCUMENT WITH LONG COMMENTS
8487273: MERGE LABEL REQUEST ON TOP OF 11.1.0.7 FOR BUGS 8214576 7258928
The database requires some initialization parameters to have a minimum value as specified in the following table:
Table 2-1 Minimum Values of Oracle Database Initialization Parameters
Property | Value |
---|---|
compatible |
10.2.0.0.0 Notes: Do not modify this parameter if it is already set to a higher value. Ensure that the value of this parameter is in the form |
db_block_size |
Either this parameter must be set to 8192, or one of the following parameters have to be set:
The db_cache_size parameter specifies in bytes of the cache of standard block size buffers. Oracle Beehive will use this value for the block size of its tablespaces. If any of the db_nk_cachce_size parameters are set, then Oracle Beehive will use the lowest value that is equal or greater than 8192. Refer to "Specifying Database Block Sizes" in "Oracle Database Administrator's Guide" for more information. |
java_pool_size |
50M (52428800) |
job_queue_processes |
10 |
processes |
150 |
sga_max_size |
1G |
sga_target |
500M |
streams_pool_size |
100M If your SGA (System Global Area) of the database is large (more than 3G), then set this parameter to 200M. |
undo_retention |
3600 |
Note:
You may have to increase the values of processes and sessions if you are using Oracle RAC.To see all parameters (in alphabetical order) along with their current values, use the following SQL*Plus command:
SQL> SHOW PARAMETERS
You may display parameters that contain a specified string. For example, to display all parameters having COMPATIBLE
in their names, use the following command:
SQL> SHOW PARAMETERS COMPATIBLE
Set initialization parameters with the ALTER SYSTEM
command. For example, to set the COMPATIBLE
parameter with the value 11.1.0.7.0
, use the following command:
SQL> ALTER SYSTEM SET COMPATIBLE = '11.1.0.7.0' SCOPE = SPFILE;
Note:
After setting initialization parameters, restart the database.The archive log mode of the database must be ARCHIVELOG
, which enables the archiving of the redo logs. Oracle Beehive uses Change Data Capture (CDC), which requires that the archive log mode be turned on. Use the following command to determine the database's archive log mode:
SQL> SELECT LOG_MODE FROM SYS.V$DATABASE;
To set the archive log mode to ARCHIVELOG
, use the following commands:
SQL> shutdown immediate SQL> startup mount exclusive SQL> alter database archivelog; SQL> alter database open;
Note:
By default, Change Data Capture is in asynchronous HotLog mode. In this mode, change data capture is performed asynchronously on the same database.For more information about modes of Change Data Capture, refer to "Change Sources and Modes of Change Data Capture" in Chapter 16, "Change Data Capture" in Oracle Database Data Warehousing Guide.
You may use Database Configuration Assistant (DBCA) to create and configure a database. For more information about this tool, refer to "Using DBCA to Create and Configure a Database" in Chapter 2, "Installing Oracle and Building the Database" in Oracle Database 2 Day DBA.
Tip:
When creating a database for Oracle Beehive with DBCA, you may choose any template. With DBCA, you may also turn on the archive log mode (by selecting the option Enable Archiving), set the character set to AL32UTF8, and set initialization parameters listed in "Initialization Parameters".Refer to "Configuring and Installing Oracle Beehive for Oracle RAC" for pre-installation and post-installation steps required for Oracle Beehive deployments that use Oracle Real Application Clusters (RAC) databases.
You may customize the layout of Oracle Beehive tablespaces with the script <Oracle home>
/beehive/db/framework/beehive_custom_ts.sql
. You may change tablespace names, initial size of the datafile, block size, and datafile autoextend options.
By default, Oracle Beehive uses eleven tablespaces. The beehive_custom_ts.sql
script maps each tablespace to a placeholder. The Oracle Beehive Install Wizard uses these placeholders (instead of the real tablespace name) for configuration.
For example, in the beehive_custom_ts.sql
script, the tablespace BEE_DATA
is mapped to the placeholder ts_data
:
REM Tables containing transactional data DEFINE ts_data = BEE_DATA DEFINE ts_data_path = '' DEFINE ts_data_sz = 512M DEFINE ts_data_bsz = 8192 DEFINE ts_data_ae = ON DEFINE ts_data_ae_nx = 128M
You must specify the mapping between the placeholders, such as ts_data
and ts_index
, and the corresponding tablespaces. If one tablespace is to be shared across multiple placeholders, adjust the tablespace size accordingly.
This script creates a special view named bee_tablespaces
, which then is used by the install scripts. You may drop the view after a successful installation. Since the tablespace mapping information can be reused for subsequent product reinstallations, the installation script does not drop it.
The script also produces an output file with the actual CREATE TABLESPACE commands. The output file name is create_beehive_tablespaces.sql
After making your changes to the script, run it with SYS privileges.
If you want to use a database that uses raw storage, customize the script <Oracle home>
/beehive/db/framework/beehive_custom_ts.sql
. Refer to "Customizing Oracle Beehive Tablespaces Layout" for more information about this script.
To use a raw device (which you have created and properly configured) for a particular tablespace, specify the full path of its datafile in the appropriate variable. For example, if you want to use a raw device for the BEE_DATA
tablespace, specify the full path of its datafile in the ts_data_path
variable.
By default, Oracle Beehive uses eleven tablespaces. The beehive_custom_ts.sql
script maps each tablespace to a placeholder. The Oracle Beehive Install Wizard uses these placeholders (instead of the real tablespace name) for configuration.
To use a raw device (which you have created and properly configured) for a particular tablespace, specify the full path of its datafile in the appropriate variable. For example, if you want to use a raw device for the BEE_DATA
tablespace, specify the full path of its datafile in the ts_data_path
variable.
In addition, set the variable APPEND_FILE_NAME
to N
.
After making your changes to the script, run it with SYS privileges.
After you have successfully installed Oracle Beehive, refer to "Oracle Beehive Database Post-Installation Procedures" for tuning recommendations to perform on your Oracle Beehive database.