sg242019
118
ORACLE Cluster POWERsolution Guide
ITSO Redbook Evaluation
Oracle Cluster POWERsolution Guide
SG24-2019-00
Your feedback is very important to help us maintain the quality of ITSO redbooks. Please complete this
questionnaire and return it using one of the following methods:
• Use the online evaluation form found at http://www.redbooks.com
• Fax this form to: USA International Access Code + 1 914 432 8264
• Send your comments in an Internet note to [email protected]
Please rate your overall satisfaction with this book using the scale:
(1 = very good, 2 = good, 3 = average, 4 = poor, 5 = very poor)
Overall Satisfaction
__________
Please answer the following questions:
Was this redbook published in time for your needs?
Yes___ No___
If no, please explain:
What other redbooks would you like to see published?
Comments/Suggestions:
(THANK YOU FOR YOUR FEEDBACK!)
117
temporary segment 9
testing the Oracle server database
81
U
user tablespace
8
W
worksheets
116
89
ORACLE Cluster POWERsolution Guide
Index
L
Symbols
/etc/oratab 98, 102
/etc/services 101
/home/oracle/.profile 97, 101
/home/oracle/tns_admin/listener.ora 98, 103
/home/oracle/tns_admin/tnsnames.ora 104
A
abbreviations 113
adapter configuration 42
archived redo log 9
archiver (ARCH) process 14
107
mixed database 16
Multi-Threaded Server 66
multithreaded server 11
mutual takeover 32
mutual takeover configuration
32
N
42
O
C
cascading cluster 32
checkpoint (CKPT) 13
client/server architecture
context area 3
control file 9
control files 67
cursor 3
4
OLTP database 16
online redo log 9
Oracle configuration files for cascading resources
Oracle user 49
Oracle user environment 51
ORACLE_HOME 57
ORACLE_SID variable 61
ORACLE_TERM 39
101
P
D
data buffer cache 10
data dictionary 7
data files 7
data segment 9
database domain name 67
database writer (DBWR) 13
DBA group 48
decision support database 16
dedicated serve 11
development database 16
disk space requirements 19
E
errors 85
extent 8
parallel server lock processes (LCKn)
parsing 3
PGA 10
process monitor (PMON) 13
Program Global Area 10
14
R
recover (RECO) process 14
redo log buffer 10
redo log files 9
rollback segment 8
rotating Ccuster 31
rotating standby configuration
30
S
H
hot standby configuration
I
in doubt transaction
index segment 9
init.ora file 10
instance 11
40
M
network type
B
bibliography
listener process 84
log writer (LGWR) 13
logical volumes for Oracle 7 storage
14
31
segment 8
SGA 10
shared pool 10
sizing 19
SQL 2
SQL language 2
system monitor (SMON)
system tablespace 8
13
T
tablespaces 8, 67
TCP service port 65
© Copyright IBM Corp. 1997
115
114
ORACLE Cluster POWERsolution Guide
List of Abbreviations
ANSI
American National Standards Institute
APA
All Points Addressable
ARCH
Archiver Process
ASCII
American Standard Code for Information Interchange
AS/400
Application System/400
CASE
Computer Assisted Software Engineering
CKPT
Checkpoint
DBWR
Database Writer
DNS
Domain Name Server
DSMIT
Distributed System Management Interface Tool
GODM
Global Object Data Manager
HACMP
High Availability Cluster Multi-Processing
HANFS
High Availability Network File System
IBM
International Business Machines Corporation
ITSO
International Technical Support Organization
LGWR
Log Writer
LOCK
Parallel Server Lock Processor
ODM
Object Data Manager
OLTP
On Line Transaction Processing
PGA
Program Global Area
PMON
Process Monitor
PROFS
Professional Office System
RECO
Recover Process
SGA
System Global Area or Shared Global Area
SMIT
System Management Interface Tool
SMON
System Monitor
VGDA
Volume Group Descriptor Area
VGSA
Volume Group Status Area
113
112
ORACLE Cluster POWERsolution Guide
IBM Redbook Order Form
Please send me the following:
Title
Order Number
First name
Quantity
Last name
Company
Address
City
Postal code
Country
Telephone number
Telefax number
VAT number
Card issued to
Signature
Invoice to customer number
Credit card number
Credit card expiration date
We accept American Express, Diners, Eurocard, Master Card, and Visa. Payment by credit card not
available in all countries. Signature mandatory for credit card payment.
DO NOT SEND CREDIT CARD INFORMATION OVER THE INTERNET.
111
How Customers Can Get ITSO Redbooks
Customers may request ITSO deliverables (redbooks, BookManager BOOKs, and CD-ROMs) and information about
redbooks, workshops, and residencies in the following ways:
• Online Orders (Do not send credit card information over the Internet) – send orders to:
In United States
In Canada
Outside North America
IBMMAIL
usib6fpl at ibmmail
caibmbkz at ibmmail
dkibmbsh at ibmmail
Internet
[email protected]
[email protected]
[email protected]
• Telephone orders
United States (toll free)
Canada (toll free)
1-800-879-2755
1-800-IBM-4YOU
Outside Notrh America
(+45) 4810-1320 - Danish
(+45) 4810-1420 - Dutch
(+45) 4810-1540 - English
(+45) 4810-1670 - Finnish
(+45) 4810-1220 - French
(long distance charges apply)
(+45) 4810-1020 - German
(+45) 4810-1620 - Italian
(+45) 4810-1270 - Norwegian
(+45) 4810-1120 - Spanish
(+45) 4810-1170 - Swedish
• Mail Orders – send orders to:
IBM Publications
Publications Customer Support
P.O. Box 29570
Raleigh, NC 27626-0570
USA
IBM Publications
144-4th Avenue, S.W.
Calgary, Alberta T2P 3N5
Canada
IBM Direct Services
Sortemosevej 21
DK-3450 Allerød
Denmark
• Fax – send orders to:
United States (toll free)
Canada
Outside Notrh America
1-800-445-9269
1-800-267-4455
(+45) 48 14 2207
(long distance charge)
• 1-800-IBM-4FAX (United States) or (+1) 408 256 5422 (Outside USA) – ask for:
Index # 4421 Abstracts of new redbooks
Index # 4422 IBM redbooks
Index # 4420 Redbooks for last six months
• Direct Services – send note to [email protected]
• On the World Wide Web
Redbooks Home Page
IBM Direct Publications Catalog
http://www.redbooks.ibm.com
http://www.elink.ibmlink.ibm.com/pbl/pbl
• Internet Listserver
With an Internet E-mail address, anyone can subscribe to an IBM Announcement Listserver. To initiate the
service, send an E-mail note to [email protected] with the keyword subscribe in the body of
the note (leave the subject line blank).
110
ORACLE Cluster POWERsolution Guide
How To Get ITSO Redbooks
This section explains how both customers and IBM employees can find out about ITSO redbooks, CD-ROMs,
workshops, and residencies. A form for ordering books and CD-ROMs is also provided.
This information was current at the time of publication, but is continually subject to change. The latest information
may be found at URL http://www.redbooks.ibm.com.
How IBM Employees Can Get ITSO Redbooks
Employees may request ITSO deliverables (redbooks, BookManager BOOKs, and CD-ROMs) and information about
redbooks, workshops, and residencies in the following ways:
• PUBORDER – to order hardcopies in United States
• GOPHER link to the Internet – type GOPHER.WTSCPOK.ITSO.IBM.COM
• Tools disks
To get LIST3820s of redbooks, type one of the following commands:
TOOLS SENDTO EHONE4 TOOLS2 REDPRINT GET SG24xxxx PACKAGE
TOOLS SENDTO CANVM2 TOOLS REDPRINT GET SG24xxxx PACKAGE (Canadian users only)
To get lists of redbooks:
TOOLS SENDTO USDIST MKTTOOLS MKTTOOLS GET ITSOCAT TXT
TOOLS SENDTO USDIST MKTTOOLS MKTTOOLS GET LISTSERV PACKAGE
To register for information on workshops, residencies, and redbooks:
TOOLS SENDTO WTSCPOK TOOLS ZDISK GET ITSOREGI 1996
For a list of product area specialists in the ITSO:
TOOLS SENDTO WTSCPOK TOOLS ZDISK GET ORGCARD PACKAGE
• Redbooks Home Page on the World Wide Web
http://w3.itso.ibm.com/redbooks
• IBM Direct Publications Catalog on the World Wide Web
http://www.elink.ibmlink.ibm.com/pbl/pbl
IBM employees may obtain LIST3820s of redbooks from this page.
• REDBOOKS category on INEWS
• Online – send orders to: USIB6FPL at IBMMAIL or DKIBMBSH at IBMMAIL
• Internet Listserver
With an Internet E-mail address, anyone can subscribe to an IBM Announcement Listserver. To initiate the
service, send an E-mail note to [email protected] with the keyword subscribe in the body of
the note (leave the subject line blank). A category form and detailed instructions will be sent to you.
109
108
ORACLE Cluster POWERsolution Guide
Appendix E. Related Publications
The publications listed in this section are considered particularly suitable for a
more detailed discussion of the topics covered in this redbook.
E.1 International Technical Support Organization Publications
For information on ordering these ITSO publications see “How To Get ITSO
Redbooks” on page 109.
• High Availability on the RISC System/6000 Family, SG24-4551
• An HACMP Cookbook, SG24-4553
E.2 Redbooks on CD-ROMs
Redbooks are also available on CD-ROMs. Order a subscription and receive
updates 2-4 times a year at significant savings.
E.3 Other Publications
These publications are also relevant as further information sources:
• HACMP for AIX, Version 4.2.1: Concepts and Facilities, SC23-1938
• HACMP for AIX, Version 4.2.1: Planning Guide, SC23-1939
• HACMP for AIX, Version 4.2.1: Installation Guide, SC23-1940
• HACMP for AIX, Version 4.2.1: Administration Guide, SC23-1941
• HACMP for AIX, Version 4.2.1: Troubleshooting Guide, SC23-1942
• HACMP for AIX, Version 4.2.1: Programming Locking Applications,
SC23-1943
• HACMP for AIX, Version 4.2.1: Programming Client Applications, SC23-1944
• HACMP for AIX, Version 4.2.1: Master Index and Glossary, SC23-1945
• Official Oracle product documentation, shipped with the database product.
Related Publications
107
to the implementation of the PTF when it becomes available to each customer
according to the normal IBM PTF distribution process.
The following terms, which are denoted by an asterisk (*) in this publication, are
trademarks of the International Business Machines Corporation in the United
States and/or other countries:
AIX ®
IBM ®
RISC System/6000 ®
AS/400 ®
PROFS ®
The following terms are trademarks of other company:
UNIX is a registered trademarks in the United States and other countries licensed
exclusively through X/Open Company Limited.
Other company, product, and service names may be trademarks or service marks
of others.
106
ORACLE Cluster POWERsolution Guide
Appendix D. Special Notices
This publication is intended to help network administrators to install and set up
highly available Oracle databases. The information in this publication is not
intended as the specification of any programming interfaces that are provided by
HACMP for AIX or for the Oracle database products. See the PUBLICATIONS
section of the IBM Programming Announcement for HACMP for AIX for more
information about what publications are considered to be product documentation.
References in this publication to IBM products, programs or services do not imply
that IBM intends to make these available in all countries in which IBM operates.
Any reference to an IBM product, program, or service is not intended to state or
imply that only IBM's product, program, or service may be used. Any functionally
equivalent program that does not infringe any of IBM's intellectual property rights
may be used instead of the IBM product, program or service.
Information in this book was developed in conjunction with use of the equipment
specified, and is limited in application to those specific hardware and software
products and levels.
IBM may have patents or pending patent applications covering subject matter in
this document. The furnishing of this document does not give you any license to
these patents. You can send license inquiries, in writing, to the IBM Director of
Licensing, IBM Corporation, 500 Columbus Avenue, Thornwood, NY 10594 USA.
Licensees of this program who wish to have information about it for the purpose
of enabling: (i) the exchange of information between independently created
programs and other programs (including this one) and (ii) the mutual use of the
information which has been exchanged, should contact IBM Corporation, Dept.
600A, Mail Drop 1329, Somers, NY 10589 USA.
Such information may be available, subject to appropriate terms and conditions,
including in some cases, payment of a fee.
The information contained in this document has not been submitted to any formal
IBM test and is distributed AS IS. The information about non-IBM ("vendor")
products in this manual has been supplied by the vendor and IBM assumes no
responsibility for its accuracy or completeness. The use of this information or the
implementation of any of these techniques is a customer responsibility and
depends on the customer's ability to evaluate and integrate them into the
customer's operational environment. While each item may have been reviewed by
IBM for accuracy in a specific situation, there is no guarantee that the same or
similar results will be obtained elsewhere. Customers attempting to adapt these
techniques to their own environments do so at their own risk.
Any performance data contained in this document was determined in a controlled
environment, and therefore, the results that may be obtained in other operating
environments may vary significantly. Users of this document should verify the
applicable data for their specific environment.
Reference to PTF numbers that have not been released through the normal
distribution process does not imply general availability. The purpose of including
these reference numbers is to alert IBM customers to specific information relative
© Copyright IBM Corp. 1997
105
This is the listener.ora file for node 2.
Figure 24. listener.ora File for Mutual Takeover Resources, Node 2
C.1.5 File: //node1_scripts/tnsnames.ora
The tnsnames.ora file is the same on both nodes.
Figure 25. tnsnames.ora File for Mutual Takeover Resources
104
ORACLE Cluster POWERsolution Guide
When there has a been a failover, we use a file called oratabD12. This contains
both database entries, as shown below.
.
D1:/home/oracle/app/oracle/product/7.3.2:Y
D2:/home/oracle/app/oracle/product/7.3.2:Y
Figure 22. oratab Entries for Both Databases Running on One Node
We copied this file into /etc/oratab. Make sure the file is copied with the oracle
user ID as its owner.
C.1.4 File: /node1_scripts/listener.ora
The figure below shows a listener.ora file for a system running more than one
database. We had two databases and thus created a Listener capable of working
with each database separately or with both databases simultaneously.
This is the listener.ora file for node 1.
Figure 23. listener.ora File for Mutual Takeover Resources, Node 1
Oracle Configuration Files for Mutual Takeover Resources
103
This is the /home/oracle/.profile file for node 2.
Figure 19. /home/oracle/.profile File on Second Node
The only difference we had between the Oracle user .profiles was the different
ORACLE_SID for each database.
You will probably need to change the following variable in the .profile files above:
•ORACLE_TERM
C.1.3 File: /etc/oratab
Below is the /etc/oratab file for nodes in a Mutual Takeover configuration.
A copy of the relevant oratab file is kept in our scripts directory. At cluster startup,
we issue:
cp -p /node1_scripts/oratabD1 /etc/oratab
This is the oratab entry for node 1.
Figure 20. Node 1 oratab Entry for the Database
This is the oratab entry for node 2. We no longer show the comments contained
in the file.
D2:/home/oracle/app/oracle/product/7.3.2:Y
Figure 21. Node 2 oratab Entry for the Database
102
ORACLE Cluster POWERsolution Guide
Appendix C. Oracle Configuration Files for Mutual Takeover Resources
This Appendix contains output from the sample cluster that was installed for an
Oracle Nonparallel Server database in a Mutual Takeover configuration.
C.1 Customized Files
The files are identical on all nodes unless otherwise stated. Where they are
different, we show both sets of files. Generally, we only show one set of files.
C.1.1 File: /etc/services
An entry for the Oracle lisd12 process with port 1521 needs to be added to the
/etc/services file. The name in the first column has to be the same as the name
used for the Listener in the listener.ora file.
We only had one Listener for all of the databases we were running. You may
decide to use more than one Listener, in which case, you will need a line for each
Listener with a different port number for each one.
.
lisD12 1521/tcp
# oracle listener
Figure 17. Entry Needed in /etc/services File for the Listener
C.1.2 File: /home/oracle/.profile
This is the /home/oracle/.profile file for node 1.
Figure 18. /home/oracle/.profile File on First Node
Oracle Configuration Files for Mutual Takeover Resources
101
100
ORACLE Cluster POWERsolution Guide
Figure 15. listener.ora File for Rotating or Hot Standby Resources
B.1.5 File: /node1_scripts/tnsnames.ora
This is the /home/oracle/tns_admin/tsnames.ora file for Rotating or Hot Standby
resources.
Figure 16. tnsnames.ora File for Rotating or Hot Standby Resources
Configuration Files for Rotating & Hot Standby Resources
99
B.1.3 File: /etc/oratab
Below is the /etc/oratab file for nodes in a Rotating or Hot Standby configuration.
.
Figure 14. /etc/oratab File for Rotating or Hot Standby Resources.
B.1.4 File: //node1_scripts/listener.ora
The figure below shows a Listener .ora file for a system running a single
database.
98
ORACLE Cluster POWERsolution Guide
Appendix B. Configuration Files for Rotating & Hot Standby Resources
This Appendix contains output from the sample cluster that was installed for an
Oracle Nonparallel Server Database in a Rotating or Hot Standby configuration.
These files are all updated correctly during the Quick Install.
B.1 Customized Files
The files are identical on all nodes unless otherwise stated. Where they are
different, we show both sets of files. Generally, we only show one set of files.
B.1.1 File: /etc/services
An entry for the Oracle Listener process with port 1521 needs to be added to the
/etc/services file. The name in the first column has to be the same as the name
used for the Listener in the listener.ora file..
lisD1
1521/tcp
# oracle listener
Figure 12. Entry Needed in /etc/services for the Listener
B.1.2 File: /home/oracle/.profile
For both nodes, this file is the same; so we only show you the Oracle user’s
.profile for one of the nodes..
Figure 13. /home/oracle/.profile File
You will probably need to change the following variable in the .profile above for
your own installation:
•ORACLE_TERM
Configuration Files for Rotating & Hot Standby Resources
97
96
ORACLE Cluster POWERsolution Guide
What terminal emulation do you choose?
What terminal emulation do you choose?
The following tables represent the raw logical volumes necessary for Oracle 7
database storage and scripts. The logical volumes for the database on node 1
and node 2 are in Table 28. Because we only require one database in a Hot
Standby configuration, the logical volumes on node2 are exactly the same as
those on node1. The names of these logical volumes are set. The numbers
represent the recomended sizes of each file. Please fill in the desired sizes.
Table 28. Database Logical Volumes for Both Nodes in Hot Standby Configuration
Name
LV Name
Default Size
Control File 1
node1_c1lv
12 MB
Control File 2
node1_c2lv
12 MB
Control File 3
node1_c3lv
12 MB
System Tablespace
node1_dblv
80 MB
Logfile 1
node1_log1lv
12 MB
Logfile 2
node1_log2lv
12 MB
Logfile 3
node1_log3lv
12 MB
Rollback Tablespace
node1_log1lv
20 MB
Temp Tablespace
node1_templv
12 MB
User Tablespace
node1_userlv
12 MB
Tools Tablespace
node1_toollv
40 MB
Your Size
Quick Installation Worksheets
95
A.3 Hot Standby Worksheet
This worksheet is to help you in gathering the necessary information to smoothly
complete the install of the Oracle7 Database on a Highlay Available Cluster using
the Quick Install Program. This worksheet is for the Hot Standby Configuration.
Please write the information in the areas provided, and keep it with you during the
installation and setup process.
Installation Information:
The following tables hold the adapter information for the cluster. In the Hot
Standby configuration each cluster has a service and standby adapter, but node
2 has no boot adapter. Please fill in the labels and IP addresses for each adapter.
As an example, the standby adapter for node 1 in our sample configuration would
have a label of node1_standby and an IP address of 10.0.2.1.
Table 26. Node1 Adapters for Hot Standby Configuration
Node 1 Adapters
Adapter Label
Adapter Address
Boot Adapter
Service Adapter
Standby Adapter
Table 27. Node2 Adapters for Hot Standby Configuration
Node 2 Adapters
Adapter Label
Adapter Address
Boot Adapter
Service Adapter
Standby Adapter
You need to find an available Group ID and User ID for both node1 and node2.
These ID #’s must be the same on both machines. The defaults used in our
sample configuration are Group #201, and User #210. To find this information,
enter smit group or smit user, and choose the list option.
Available Group ID #
Available User ID #
Pick a terminal emulation that is supported by Oracle, and accepted by your
hardware. If you have an incorrect terminal emulation, you will not be able to
install the database. The following are supported by Oracle:
3151, 386*, Q303*, Q310*, ansi, dec, hft*, hp, iris, lft, sun, vt100, vt220,wy50,
wy150, xlft, xsun, xsun50 (* Multiple versions of these emulations are available.)
94
ORACLE Cluster POWERsolution Guide
What terminal emulation do you choose?
The following tables represent the raw logical volumes necessary for Oracle 7
database storage and scripts. The logical volumes for the database on node1 and
node2 are in Table 25 on page 93. Because we only require one database in a
Rotating configuration, the logical volumes on node2 are exactly the same as
those on node1. The names of these logical volumes are set. The numbers
represent the recomended sizes of each file. Please fill in the desired sizes.
Table 25. Database Logical Volumes for Both Nodes in Rotating Configuration
Name
LV Name
Default Size
Control File 1
node1_c1lv
12 MB
Control File 2
node1_c2lv
12 MB
Control File 3
node1_c3lv
12 MB
System Tablespace
node1_dblv
80 MB
Logfile 1
node1_log1lv
12 MB
Logfile 2
node1_log2lv
12 MB
Logfile 3
node1_log3lv
12 MB
Rollback Tablespace
node1_log1lv
20 MB
Temp Tablespace
node1_templv
12 MB
User Tablespace
node1_userlv
12 MB
Tools Tablespace
node1_toollv
40 MB
Your Size
Quick Installation Worksheets
93
A.2 Rotating Standby Worksheet
This worksheet is to help you in gathering the necessary information to smoothly
complete the install of the Oracle7 Database on a Highlay Available Cluster using
the Quick Install Program. This worksheet is for the Rotating Configuration.
Please write the information in the areas provided, and keep it with you during the
installation and setup process.
Installation Information:
The following tables hold the adapter information for the cluster. In the Rotating
configuration each cluster has a standby and boot adapter, but both share a
service adapter. Please fill in the labels and IP addresses for each adapter. As an
example, the shared adapter for node 1 and node 2 in our sample configuration
would have a label of node1 and an IP address of 10.0.1.1.
Table 23. Node 1 Adapters for Rotating Configuration
Node 1 Adapters
Adapter Label
Adapter Address
Boot Adapter
Service Adapter
Standby Adapter
Table 24. Node 2 Adapters for Rotating Configuration
Node 2 Adapters
Adapter Label
Adapter Address
Node1 Shared Adapter
Node1 Shared Adapter
Boot Adapter
Service Adapter
Standby Adapter
You need to find an available Group ID and User ID for both node1 and node2.
These ID #’s must be the same on both machines. The defaults used in our
sample configuration are Group #201, and User #210. To find this information,
enter smit group or smit user, and choose the list option.
Available Group ID #
Available User ID #
Pick a terminal emulation that is supported by Oracle, and accepted by your
hardware. If you have an incorrect terminal emulation, you will not be able to
install the database. The following are supported by Oracle:
3151, 386*, Q303*, Q310*, ansi, dec, hft*, hp, iris, lft, sun, vt100, vt220,wy50,
wy150, xlft, xsun, xsun50 (* Multiple versions of these emulations are available.)
92
ORACLE Cluster POWERsolution Guide
Name
LV Name
Default Size
Temp Tablespace
node2_templv
12 MB
User Tablespace
node2_userlv
12 MB
Tools Tablespace
node2_toollv
40 MB
Your Size
Quick Installation Worksheets
91
Pick a terminal emulation that is supported by Oracle and accepted by your
hardware. If you have an incorrect terminal emulation, you will not be able to
install the database. The following are supported by Oracle:
3151, 386*, Q303*, Q310*, ansi, dec, hft*, hp, iris, lft, sun, vt100, vt220,wy50,
wy150, xlft, xsun, xsun50 (*Multiple versions of these emulations are available.)
What terminal emulation do you choose?
The following tables represent the raw logical volumes necessary for Oracle 7
database storage and scripts. The logical volumes for the database on node1 are
in Table 21 on page 90. The logical volumes for the database on node2 are in
Table 22 on page 90. The names of these logical volumes are set. The numbers
represent the recomended sizes of each file. Please fill in the desired sizes.
Table 21. Database Logical Volumes (Node 1)
Name
LV Name
Default Size
Control File 1
node1_c1lv
12 MB
Control File 2
node1_c2lv
12 MB
Control File 3
node1_c3lv
12 MB
System Tablespace
node1_dblv
80 MB
Logfile 1
node1_log1lv
12 MB
Logfile 2
node1_log2lv
12 MB
Logfile 3
node1_log3lv
12 MB
Rollback Tablespace
node1_log1lv
20 MB
Temp Tablespace
node1_templv
12 MB
User Tablespace
node1_userlv
12 MB
Tools Tablespace
node1_toollv
40 MB
Your Size
Table 22. Database Logical Volumes (Node 2)
90
Name
LV Name
Default Size
Your Size
Control File 1
node2_c1lv
12 MB
12M
Control File 2
node2_c2lv
12 MB
12M
Control File 3
node2_c3lv
12 MB
12M
System Tablespace
node2_dblv
80 MB
Logfile 1
node2_log1lv
12 MB
Logfile 2
node2_log2lv
12 MB
Logfile 3
node2_log3lv
12 MB
Rollback Tablespace
node2_log1lv
20 MB
ORACLE Cluster POWERsolution Guide
Appendix A. Quick Installation Worksheets
This Appendix provides worksheets to fill out in preparation for running the Quick
Install Program.
A.1 Mutual Takeover Worksheet
This worksheet is to help you in gathering the necessary information to smoothly
complete the install of the Oracle 7 database on a highly available cluster using
the Quick Install Program. This worksheet is for the Mutual Takeover
configuration. Please write the information in the areas provided, and keep it with
you during the installation and setup process.
Installation Information:
The following tables hold the adapter information for the cluster. In the Mutual
Takeover configuration, each cluster has a service, standby and boot adapter.
Please fill in the labels and IP addresses for each adapter. As an example, the
boot adapter for node 1 in our sample configuration would have a label of
node1_boot and an IP address of 10.0.1.11.
Table 19. Node1 Adapters for Mutual Takeover Configuration
Node1 Adapters
Adapter Label
Adapter Address
Boot Adapter
Service Adapter
Standby Adapter
Table 20. Node2 Adapters for Mutual Takeover Configuration
Node2 Adapters
Adapter Label
Adapter Address
Boot Adapter
Service Adapter
Standby Adapter
You need to find an available Group ID and User ID for both node1 and node2.
These ID #’s must be the same on both machines. The defaults used in our
sample configuration are Group #201, and User #210. To find this information,
enter smit group or smit user, and choose the list option.
Available Group ID #
Available User ID #
Quick Installation Worksheets
89
88
ORACLE Cluster POWERsolution Guide
• Check to see that the service name is mapped to a connect descriptor in the
tnsnames.ora file and add or correct it if necessary.
• Make sure there are no syntax errors in the tnsnames.ora file; check to see if it
has unmatched parentheses or stray characters.
TNS-12203: TNS: unable to connect to destination
This error indicates that the client is not able to find the desired database. If you
get this error message, do the following:
• Be sure you have correctly entered the service name you want to reach.
• Look at the tnsnames.ora file to see that the ADDRESS parameters in the
connect descriptor for the service name are correct.
• Check and verify if the Listener on the remote node is up and listening. You
can verify this by typing lsnrctl status <listener_name>:
TNS-12545: TNS: name lookup failure
This error occurs when the Listener on the remote node cannot be contacted.
The ADDRESS in the tnsnames.ora file or the Listener.ora file may be incorrect.
This message also appears if the Listener on the remote node has not been
started.
Note: The error messages in SQL*Net or client-to-server connection errors are
usually caused by errors in the tnsnames.ora or Listener.ora files. Therefore,
these files must be rechecked when you receive any of these errors.
For more information about all SQL*Net error messages, please refer to the
Oracle Network Products Message Manual.
8.2 Recommendations
We recommend strongly that you put the control files, redo log files, and all the
database object files, such as system tablespace, rollback tablespace, temporary
tablespace, and user tablespace, in the shared external disk in the HACMP
environment.
The reason is that the location of these files is one of the critical points that
determines whether or not your database is highly available.
We also recommend that before you do all the installation tasks, carefully plan
your database and prepare your system. For example, make sure your cluster
functions properly without the database application. Carefully planning your
database means planning all the names and variables for your database, the size
of your database, and making sure you have the prerequisite hardware and
software. For parallel databases, you also have to create all the raw logical
volumes needed for all of the database object files prior to the installation
process.
Common Errors, Problems and Recommendations
87
For example, you may see the following error message in your Installer menu
while creating the database objects:
ORA-00210: cannot open control file /oradatafs2/ora_ctl/ctl1.ora
From this, you know there is something wrong with your control file. Maybe the
full pathname you specified is not correct, or your directory is inaccessible. You
need to look at your directory permissions.
For complete information about error messages (cause and action), please refer
to the Oracle 7 Server Messages and Codes Manual.
8.1.1.2 Postinstallation-Stage Common Errors and Problems
Remember that every time you finish an installation, you must do the
postinstallation tasks. If you forget to complete these tasks, an error situation will
result, and you won’t be able to run your installed products.
You need to run root.sh after you finish the installation.
# ./root.sh
You might find that the program always says:
Unmatched variables for ORACLE_SID or ORACLE_HOME.
If this occurs, you have to do the following:
# ORACLE_SID= <your oracle_sid> ; export ORACLE_SID
# ORACLE_HOME= <your oracle_home> ; export ORACLE_HOME
and as root, and run the root.sh again.
8.1.2 SQL*Net Errors and Problems
When you try to run the Listener on the server by typing:
$ lsnrctl start
you might receive this message:
TNS-01151: Missing listener name, lisD1 or lisD12, in LISTENER.ORA
The message above means than you do not have a Listener name in your
Listener.ora file, or that you have a Listener name, but did not name it LISTENER
or the default name. You have to check your Listener.ora file and see if you have
specified your Listener name (see the example in the appropriate Appendix to
your configuration). If you have, you should run the Listener by typing the Listener
name after the lsnrctl start command.
The other common error message received when you run the Listener is shown
below:
TNS-12154: TNS: could not resolve service name
This error points to a problem related to the tnsnames.ora file. The service name
specified cannot be located in the tnsnames.ora file. If you get this error
message, do the following:
• Verify that your tnsnames.ora exists and is in the proper directory. Please refer
to the previous chapter sections that discuss postinstallation tasks on the
server and postinstallation tasks on the client.
86
ORACLE Cluster POWERsolution Guide
Chapter 8. Common Errors, Problems and Recommendations
This chapter discusses common Oracle error messages that you might encounter
and problems you might have when you do the installation and setup. Also, we
will give some recommendations concerning the highly available database.
All of the discussions here are based on our experiences while preparing and
doing all the tasks described in previous chapters. Therefore, this chapter is not
intended to explain all errors and problems in Oracle. You might find errors or
problems that differ from this chapter when you do your installation. However, if
you run into some of the same problems that we did, perhaps this chapter will be
helpful to you.
8.1 Common Errors and Problems
This section describes several common errors that occur during the installation
and during testing of database access from the client to the server.
For complete information about error messages, please refer to the Oracle 7
Server Messages and Codes Manual and to the SQL*Net Administration Guide.
Another good reference is the Appendix B discussion about “Troubleshooting
Your Installer Session” in the Oracle 7 for AIX-Based Installation and
Configuration Guide for handling problems with the Installer.
8.1.1 Errors and Problems During the Installation Process
When you want to install the Oracle products, always remember the three stages
of the process: preinstallation, installation, and postinstallation.
Failure to complete one of these three tasks will end in error, and you will not be
able to run your database server and other Oracle products.
8.1.1.1 Installation-Stage Common Errors and Problems
This table shows the common errors that might happen while installing and helps
you to identify the possible reasons for the problem.
Table 18. Common Errors During the Installation Stage
Stage of Installation
Installing the Installer
Common Error
Initial Installer prompts
Loading the product file
System requirement does not meet
environment setup
Environment setup
Not enough space in staging area
Installing the product files
File permissions problem
Not enough space in $ORACLE_HOME
System requirement not met
Relinking product executables
Creating database objects
File permissions problem
Not enough space
System requirement not met
System configuration (such as shared
memory)
Database access problems
Common Errors, Problems and Recommendations
85
Figure 11. Oracle Processes Running on Node with Database
The Listener process is lisD1 (PID 10670).
To test a failover, simply run the Stop Cluster Services SMIT screen and select
the graceful with takeover option, or power-off the machine running the
database. With Rotating or Hot Standby, this will be one of the nodes in the
cluster. With Mutual Takeover, you can power-off either machine because each is
serving its own database. On the machine in the cluster that is still running,
execute the following command to watch the takeover activity:
tail -f /tmp/hacmp.out
84
ORACLE Cluster POWERsolution Guide
7.4 Testing Cluster Synchronization
Testing the HACMP portion of the install can be done with HACMP’s cluster
verification utility. To get to this utility, enter smit hacmp from the command prompt.
Select Cluster Configuration. From the next menu, select Cluster Verification.
At the prompt, just press Enter until the process begins. Run this tool on both
nodes of the cluster.
7.5 HACMP Cluster Services
When are ready to begin failover testing, you must first start the HACMP services
on both nodes. One node at a time, from the command prompt, enter smit hacmp.
Select Cluster Services. The next menu will allow you to either start, stop, or
check on the cluster services.
After selecting Start Cluster Services, press Enter until the process begins.
When the process is finished, and you get the OK prompt, you can exit out of
SMIT by pressing the F10 key.
To see the progress of the cluster, from the command prompt, type:
# tail -f /tmp/hacmp.out
This will monitor the progress of the startup. When you want exit out of this, press
Ctrl-C.
When you are done, bring up the second node in the same way. To check if the
database is running, type:
# ps -eaf |grep ora
from the command prompt. You should see something similar to Figure 11.
Post-Installation of the Oracle Server Database
83
Mutual Takeover Configuration:
Rotating/Hot Standby Configuration:
If any of these files are missing, please copy them from the casc_scripts (for
Mutual Takeover) or rot_scripts (for Rotating/Hot Standby) directory on the
accompanying disk. The content of these files will be checked later in this
chapter.
82
ORACLE Cluster POWERsolution Guide
Chapter 7. Post-Installation of the Oracle Server Database
The Quick Install Program takes you through the post-installation tasks, updates
the control files with the appropriate variables, and moves them to the required
directories. When you are done with the Quick Install, you are ready to test the
database. However, this an outline of the files that have been manipulated from
their defaults by the Quick Install Program.
7.1 Verify Database Security and Integrity
To maintain integrity of the Oracle code, all executables in the
$ORACLE_HOME/bin directory must be writable by the Oracle owner only.
You must verify that these files have file modes 755 or 777. Also, to maintain
discretionary access to data, all databases, redo log and control files that may be
put in the external disks must be readable by the Oracle owner and group only,
including the root user for highly available purposes.
7.2 Create Oracle Server User Logins
For each Oracle user, create a login with the following properties:
Login name
The user’s user name
User ID number
A number between 3 and 32767
Primary Group
A number between 3 and 32767, not the dba group
Home directory
/home/oracle
Login shell
/bin/ksh
7.3 Check your Filesystems
The filesystems you created on the external disks (node1_scripts in our example)
must contain the necessary control scripts and data files. Below is a copy of the
files stored in the node1_scripts directory by the Quick Install Program for the
Mutual Takeover configuration and for the Rotating/Hot Standby configurations:
Post-Installation of the Oracle Server Database
81
YOU HAVE FINISHED WITH THE QUICK INSTALL SCRIPT FOR NODE 1.
PLEASE RETURN TO THE TOP AND REPEAT FOR NODE 2. THERE IS LESS
USER INPUT IN NODE 2, AND THUS A FASTER INSTALL.
80
ORACLE Cluster POWERsolution Guide
Enter OK to get through both screens; then enter Yes at the last screen to
continue on.
6.46 Optional Product Documentation
If you choose optional products you want to install (for example, in this
configuration, we chose SQL*Plus V2), you are prompted for these by the
Installer. For complete information about the product installation prompts, see the
Oracle 7 Server for AIX-Based Installation and Configuration Guide.
Select No in the SQL Plus V2 Demo Tables and Help Facility Screens.
The Installer will do the rest of the work. This portion may take up to an hour
depending on your hardware. Near the end of the install, you will see two or three
messages built into the Installer. Read these, and then press OK to continue.
When it is done, you will need to tab to Exit to exit the Installer.
6.47 Post-Installation Program
The output of the root.sh script will be something like the following example:
Running ORACLE7 root.sh script...
The following environment variables are set as :
ORACLE_HOME= /home/oracle/app/oracle/product/7.3.2
ORACLE_SID= D1 (or D2)
Are these settings correct (Y/N)? [Y]:
These settings should be fine; so enter Y to continue.
Enter the full pathname of the local bin directory [/usr/lbin]:
This local directory was created earlier in the install process. Just press the
Enter key since the default is the correct directory.
You should get a warning message about the $ORACLE_HOME directory not
being consistent. This message is expected. When you are prompted, just
enter Yes to continue.
Depending on which products you have selected to install, messages are
displayed to alert you to the progress of the root.sh actions.
Attention: Because you log in as root, you may have to set your ORACLE_HOME,
ORACLE_SID, and ORACLE_OWNER variables by issuing the following commands on
the command line according to your environment. Below are the commands
we issued for the sample configuration:
# ORACLE_HOME= /home/oracle/app/oracle/product/7.3.2 ; export ORACLE_HOME
# ORACLE_SID= D1 ; export
Note: D1 is the ORACLE_SID for node 1, and D2 is the ORACLE_SID for node 2.
6.48 Copying and Linking Necessary Files
After the copying of necessary files, you are finished with the setup of the first
node of the cluster.
You must now shut down the machine. At the prompt enter:
# shutdown -Fr
Running the Quick Install Program
79
Tools Tablespace Data File:
Size of Tools Tablespace:
You are now at the DB Defaults screens again. This time, however, they contain
the data you just entered. Check them carefully. Make sure there are no mistakes.
78
ORACLE Cluster POWERsolution Guide
User Tablespace Data File:
Size of User Tablespace:
Running the Quick Install Program
77
Temporary Segment File:
Size of Temporary Segment File:
76
ORACLE Cluster POWERsolution Guide
Rollback Segment File:
Size of Rollback Segment File:
Running the Quick Install Program
75
Third Redo Log:
Size of Third Redo Log:
74
ORACLE Cluster POWERsolution Guide
Second Redo Log:
Size of Second Redo Log:
Running the Quick Install Program
73
First Redo Log:
Size of First Redo Log:
72
ORACLE Cluster POWERsolution Guide
System Tablespace Data File:
Size for System Tablespace:
Running the Quick Install Program
71
The third control file:
6.45.1 Name and Size Tablespaces and Logs:
The next few pages show the prompts with the answers highlighted in black. You
will be shown the DB Defaults (two screens). These have the wrong names and
sizes; so enter OK to get through both screens, and enter No at the last screen to
go back and enter the correct names and sizes.
Note: Remember that the sizes entered will be one MB less than the actual size
of the logical volume created earlier. This is to prevent any disk space problems
later.
Here we go:
70
ORACLE Cluster POWERsolution Guide
The first control file:
The second control file:
Running the Quick Install Program
69
6.44 Database Domain Name
The DB domain name is just an identifier, but to keep it consistent, use the SID
name (D1 or D2), an underscore and the word domain. For node 2 on a Mutual
Takeover configuration, use D2_domain.
6.45 Name Control Files and Tablespaces
These prompts are asking for the names of the raw logical volumes that you
created earlier. As stated before, this is where the database is stored. You should
have these names on the Quick Install worksheet as well as on a table in the
previous chapter.
68
ORACLE Cluster POWERsolution Guide
6.39 Choose Storage Type
The database storage is done on raw logical volumes.
6.40 Select the Character Set
You can choose a character set as listed from the Installer’s prompt:
The default value is US7ASCII.
6.41 Enter the Password for the SYSTEM User and the SYS User
Note: The SYSTEM and SYS users are automatically created for you when you
install Oracle Server. These users have dba privileges.
6.42 Do You Want To Set the Passwords for DBA and Operator?
The default is No, but if you want to set passwords for these two internal users,
allowing them to connect internally without passwords, then you must choose
YES.
6.43 MTS Configured and the SQL*Net Listener Automatically Started?
The Multi-Threaded Server and the Listener should be on at all times, and
therefore they should be configured at system startup. Select Yes.
Running the Quick Install Program
67
6.35 TCP Service Port
The TCP Service Port was entered into the /etc/services file earlier in the install.
It is always set to 1521; so use the backspace to delete the 8888 and enter 1521.
6.36 Passwords
Oracle 7 has many passwords, and you are asked to enter them all during this
portion of the install. The following screen is the only one we will show you, but
rest assured, there are many. Each is followed by a password-verification prompt.
Make sure these are the same, or you will get an error.
6.37 Group Names
The next two prompts ask you for the Database Administrator Group and the
OSOPER Group. For the first screen, select the radio button next to dba. For the
second, just enter dba, because this was the group we created earlier for
administrators.
6.38 More Than One Instance?
We do not want to allow more than one instance of the database; so we will enter
No, to this next prompt to keep the maximum number of instances equal to one.
66
ORACLE Cluster POWERsolution Guide
Note: For the cluster DAVE sample configuration, we only chose Oracle 7 Server
(RDBMS) 7.3, SQL*Net V2, SQL*Plus 3.3, and the TCP/IP Protocol Adapter V2.
(Not all shown in picture)
6.34 Official Hostname
You are prompted for the Official Hostname for the server. This refers to the label
of the service adapter for the machine. Remember, that Rotating and Hot
Standby configurations have the same database setup for node 1 and node 2; so
both machines will have node 1’s service adapter as an answer to this prompt.
Running the Quick Install Program
65
6.32 Online Help and Documentation
In the sample configuration, we chose not to install any online help or
documentation. The next three screens ask for the level of help and
documentation you wish to install. If you choose any, make sure you consult the
disk sizing worksheets in the previous chapter.
The next screen is for UNIX Documentation; we chose No.
The third screen is the Product Documentation Library, which is held on an
additional CD-ROM. As in the screen shown above, we selected No Products.
6.33 Products Menu
By now, you see that the Installer prompts you with the list of Oracle products that
reside on your distribution media (tape or CD).
Use the arrow keys to scroll to, and the space bar to select, one or more Oracle
products that you want to install.
After you have chosen the products, use the Tab key to navigate to the (Install)
button, and press Enter or Return to activate the installation phase.
64
ORACLE Cluster POWERsolution Guide
For this sample configuration, the value is American/English.
6.30 Relink Executables?
The Installer offers you the option to relink the executables with the following
prompt:
6.31 Location of Postinstallation File
The next screen tells you the location of the postinstallation file. This is always,
/home/oracle/app/oracle/product/7.3.2/orainst/root.sh. Enter OK to continue.
Running the Quick Install Program
63
6.27 Source Staging Area
Enter the pathname of the source staging area. The staging area is located in
/home/oracle/stage. This area will contain links to the CD-ROM.
6.28 ORACLE_SID Variable
The next screen asks for your ORACLE_SID variable, which is the identifier for the
database. This should be automatically read out of your environment
(/home/oracle/.profile) area. For node 1, the SID is D1; for node 2 in the Mutual
Takeover configuration, the SID is D2.
6.29 Specify National Language Support (NLS)
The Installer allows you to specify a language in which to receive screen
messages for those utilities that support NLS.
62
ORACLE Cluster POWERsolution Guide
6.26 Installation Source
The next two prompts ask for the install source. We will be installing from a
temporary staging area.
Note: Remember to delete the staging area when you have finished.
Running the Quick Install Program
61
6.24 Preinstallation Preparation
The Quick Install Program has run the rootpre.sh preinstallation program for you;
so select Yes.
6.25 Install Products on All Nodes?
The next prompt asks if you want to install the products on all the nodes that
comprise the cluster. Select No, because we only want to install the products on
this node.
60
ORACLE Cluster POWERsolution Guide
6.21 Create Database Objects
Select Yes in the DB Objects menu to create the database. The database
prompts will come later in the install. If you miss this step, the database creation
prompts will be skipped.
6.22 Information Prompts
Continue on through the next series of information prompts. These provide you
with information such as log locations and fixes. Just answer OK to these
screens.
6.23 OPS Install
You do not wish to install the products on all the nodes that comprise the cluster
for the Oracle 7 Parallel Server. Select No from the OPS Install prompt.
Running the Quick Install Program
59
Note: This screen has radio buttons (the areas next to the options). To click the
radio button, use the Tab key to move the highlighted cursor to the selectable
fields, the up and down arrow keys to move the cursor to the option you want, and
the Space bar to select that option.
6.19 Installation Options
At the Installation Options menu, you select Install New Product.
6.20 Entering the Mount Point
You must define the Mount Point for the software installation. This is different from
the ORACLE_HOME pathname. The ORACLE_HOME variable is set to
/home/oracle/app/oracle/product/7.3.2. The mount point is /home/oracle.
The next prompt will fill in the complete ORACLE_HOME location. Again, it should be
/home/oracle/app/oracle/product/7.3.2. If that is correct, just enter OK.
58
ORACLE Cluster POWERsolution Guide
6.17 Oracle 7 Welcome
You are now launched into the Oracle 7 Install program. There are many
prompts.
6.18 Enter Installer Activity
After you have started the Installer, you are asked to choose an Installer activity.
You can choose to Install, Upgrade, or De-Install Software.
Running the Quick Install Program
57
6.15 Linking Oracle 7 Files to Staging Area
The install files will now be loaded into a staging area. Your are prompted for the
name of the staging area for oracle linking, which was created for you earlier. At
the prompt, type: /home/oracle/stage.
NOTE: This may take up to forty-five minutes.
6.16 Running The Preinstallation Script
When the linking has been completed, the Oracle 7 preinstallation script,
rootpre.sh is run automatically. This file sets some environment variables.
56
ORACLE Cluster POWERsolution Guide
6.14 Mount CD-ROM
Make sure you have the CD-ROM marked “Oracle 7 Server” in the drive. When
you are ready, enter y to continue.
Running the Quick Install Program
55
6.13 Creation of Logical Volumes
The list of default logical volumes now appears.
You will notice that this list matches the list from your worksheet. Match up the
sizes with those you have chosen for your database (the defaults are shown). If
you have any changes, just select the line number of the logical volume you wish
to change. If you want to continue enter 12, and if you want to exit at this point,
enter 13.
When you continue on, you will see each logical volume being created one by
one. This may take some time.
When all the logical volumes have been created, the permissions will be set on all
of them.
NOTE: If you chose a Mutual Takeover Configuration, you have two databases;
you therefore must have two volume groups, and two filesystems for storage.
Therefore, you will now be taken back to the Create Volume Group prompt for a
second time through.
NOTE: On node 2, the volume groups and logical volumes will be imported from
node 1.
54
ORACLE Cluster POWERsolution Guide
continue on. If the volume group does not exist, you will be asked to create it, and
you are shown a list of available disks. Select a disk by entering its line number.
Once you have selected a disk, you are shown the disk you have already chosen
and asked if you would like to use another disk in the creation of the volume
group.
If you select no (n), the volume group will be created. If you select yes (y), you
are shown the available disks again and asked to choose. When you have
selected all the disks you need, select no, and the volume group will be created.
As the volume group is being created, the filesystem you named earlier is also
being created. This filesystem will store the Oracle 7 control scripts for the
database it is affiliated with.
If there is only one disk available, it will automatically be chosen and used to
create the volume group.
NOTE: It is a good idea, when using two or more disks to create a volume group,
to select disks attached to adapters in different slots to allow for mirroring.
Mirroring will make two copies of the logical volumes on the volume group,
making the cluster even more secure against failure. You will be prompted after
creation as to whether or not you want mirroring.
Running the Quick Install Program
53
6.11 Setting Oracle User Environment
Oracle 7 requires several variables to be properly set in the superuser’s
environment. These variables, ORACLE_HOME, ORACLE_TERM, and ORACLE_SID are set
for you. The only information you need to enter is the terminal type you found
earlier and entered on your worksheet. This type must be compatible with Oracle
7 as well as with your hardware and must allow for you to backspace or delete
text.
6.12 Creating Volume Groups
The volume group names you entered earlier must now be created if they do not
already exist. This part of the Quick Install Program will check the name of the
volume group and see if it exists. If so, it will check it for sufficient disk space and
52
ORACLE Cluster POWERsolution Guide
number you entered is not available, or an oracle user already exists on your
system.
Once this user is created, you need to set the password for the account. This
password is temporary. When you log in to the machine for the first time as
oracle, you will be prompted to change the password again.
6.10 Creating Filesystems, Directories, and Setting Permissions
This area of the Quick Install Program requires no user input. There are status
screens flashed on screen, as well as messages of success or error. The
following is occuring during this time:
• First, the /home/oracle filesystem is created. This filesystem is created on
rootvg, and it holds all the local Oracle 7 files and scripts. The volume group is
checked for sufficient disk size (300 MB) and then created. Permissions are
then set on this directory.
• Next, a local bin is created for the Oracle 7 database to store commands in.
Permissions are set on this directory (/usr/lbin).
• An entry for the Oracle Listener is created in the /etc/services file. The
Listener handles communication between the users and the database.
• Lastly, another filesystem is created to be a temporary staging area for the
Oracle 7 files. The CD-ROM files are copied and then linked to this area
(/home/oracle/stage) for installation. Rootvg is checked for sufficient space
before the filesystem is created. The proper permissions are then set on the
/home/oracle/stage directory.
This directory may be deleted after installation is complete.
Running the Quick Install Program
51
created for you. If you get an error message, the number you entered is not
available, or there is an existing dba group on your system.
6.9 Create Oracle User Account
The Oracle user (oracle) is the superuser. This account has access to all areas of
the database and permissions on all administrative functions. This user has a
home directory at /home/oracle on the system, and all the local database files are
located there. You are prompted for the user ID number. This number was found
earlier, and should be written on your worksheet. The default is 210. After you
enter this number, the user will be created. If you receive an error message, the
50
ORACLE Cluster POWERsolution Guide
6.7 Change Default Names
After the warning on space requirements, you will come to a set of menus that
allows you to change the default names for the volume groups and filesystems
that will be created on your cluster. The volume groups will float between the
nodes of the cluster during failover and will hold the filesystems where database
scripts and data reside. Follow the menu options to make changes to the default
names.
NOTE: These menus will not appear in node 2. The names have been read from
node 1.
The defaults for a Mutual Takeover configuration are:
node1vg node1_scripts
node2vg node2_scripts
The defaults for a Rotating or Hot Standby configuration are:
node1vg node1_scripts
When you have finished with the changes, enter 2 to continue on or 3 to quit the
program.
6.8 Create DBA Group
This prompt asks you to enter the available group ID number that you found on
your worksheet. This Database Administrator Group (dba) contains the database
superusers. Once you enter the number (the default is 201), the user will be
Running the Quick Install Program
49
6.6 Information Menus
You are now entering the Oracle portion of the install. This Welcome screen lays
out what will be done in this section of the Quick Install. Enter y to continue, or
type exit to leave the Installer.
The next screen is a warning to make sure that you have planned the disk space
sizing properly for installation. You should have planned this in the previous
chapter; however, if you need to leave the program to make adjustments at this
time, type exit. There are several explanations of disk sizing in this manual. If you
have already done this, enter y to continue.
48
ORACLE Cluster POWERsolution Guide
Hot Standby Node 2:
NOTE: On node 2, after the adapters are configured, you will be prompted for the
label of node 1’s boot adapter. This information is necessary for comunication
between the nodes during installation.
Running the Quick Install Program
47
Rotating Standby Node 2:
Hot Standby Node 1:
46
ORACLE Cluster POWERsolution Guide
Mutual Takeover Node 2:
Running the Quick Install Program
45
To make any changes to the label, address, mask, or slot position of the adapters,
just follow these instructions:
• To enter/change an adapter label, enter L and a number (Ex. L0, L1).
• To enter/change an IP Address, just enter A and a number (Ex. A0,A2).
• To enter/change the subnet mask for an adapter just enter M and a number (Ex
M0,M1).
• To swap the position of two adapters, just enter S followed by the numbers you
want switched (Ex. S12).
When you have filled in all the appropriate information enter C to move on.
Below, we will display the screens of node1 and node2 for all three
configurations, as they are all unique. Make sure you are using the appropriate
information for your configuration.
Mutual Takeover Node 1:
44
ORACLE Cluster POWERsolution Guide
6.3 Choose TTY Port
The available TTY lines are presented to you. This line supports the “heartbeat”
between the nodes in the cluster. Please select the appropriate choice for your
machine:
6.4 Select Network Type
Select 1 if your network is using Ethernet connections, or select 2 if your network
is using token-ring connections.
6.5 Adapter Configuration Screen
The next screen does all the adapter configuration for you. You must enter all the
appropriate information, however. You should have gathered this information
while completeing the configuration worksheet. If you have an existing network,
and are using IP addresses from that network, the Quick Install Program will
locate this information, and present it on screen. If you do not have an existing
network, thiis screen will come up with the default information. Either way, consult
your worksheet for the proper adapter information and make any changes on this
screen.
Running the Quick Install Program
43
6.2 Declaring Failover Type
You should have already chosen which failover type you wish to use for this
cluster configuration. If you want a Mutual Takeover configuration, enter 1; for
Rotating Standby, enter 2, and for Hot Standby, enter 3.
42
ORACLE Cluster POWERsolution Guide
Chapter 6. Running the Quick Install Program
This chapter outlines the steps involved in installing an Oracle database on a
highly available cluster. The Quick Install Program used here does not apply to
Oracle Parallel Server.
Now that you have the necessary information to begin installation, follow these
steps, and let the Quick Install program lead the way. The steps shown are for
node 1. Any changes for node 2 are specified along the way.
NOTE: If you choose to leave the Installer, or are forced out due to an error,
running the script called cleanup and then shutting down the node will prepare the
machine for the installation process again.
Start the Quick Install Program by running the following commands as root:
# cd /tmp/hascripts
# ./setup
6.1 Declaring Node Number
This first prompt asks you to enter the node which you are currently working on. If
you are working on node 1, enter 1. If you are working on node 2, enter 2.
Running the Quick Install Program
41
5.9 Creation of Logical Volumes
These raw logical volumes will store your database. They are created during
the Quick Install program, but you are given the opportunity to adjust the
default sizes. Check the Oracle 7 Installation Guide for a better explanation of
the logical volumes. When you have decided upon the perfect sizes, enter
them on your worksheet.
Remember, if you are using a Mutual Takeover configuration, you will have a
set of logical volumes for each database. Therefore, you must pick the sizes
for both sets of logical volumes.
Oracle Object
LV name
Oracle File Name
Suggested
LV size (MB)
# of Logical
Partitions
Control file 1
c1lv
/dev/rnode1_c1lv
N/A
N/A
Control file 2
c2lv
/dev/rnode1_c2lv
N/A
N/A
Control file 3
c3lv
/dev/rnode1_c3lv
N/A
N/A
System tablespace
dblv
/dev/rnode1_dblv
80 MB
20
Logfile 1 (instance 1)
log1lv
/dev/rnode1_log1lv
12 MB
3
Logfile 2 (instance 1)
log2lv
/dev/rnode1_log2lv
12 MB
3
Logfile 3 (instance 1)
log3lv
/dev/rnode1_log3lv
12 MB
3
Rollback tablespace
rollv
/dev/rnode1_rollv
20 MB
5
Temp tablespace
templv
/dev/rnode1_templv
12 MB
3
User tablespace
userlv
/dev/rnode1_userlv
12 MB
3
Tools tablespace
toollv
/dev/rnode1_toollv
40 MB
10
Figure 10. Logical Volumes for Oracle 7 Storage
The sizes given for the logical volumes are the recommended to load Oracle 7
and the sample database that comes with it. These are more than double the
minimum sizes laid out by Oracle. For more detailed information, please see
the Oracle7 documentation.
Note: A logical partition is equivalent to 4 MB.
40
ORACLE Cluster POWERsolution Guide
NOTE: Once you have entered these two files, you must check your adapters.
After shutting down the system, and logging back in, type host followed by the
adapter name. In Figure 9, we tested node1_standby:
Figure 9. Adapter Test on node1_standby
You need to test every adapter in this way, and check that the IP address returned
is the address expected. If not, you have a name resolution conflict on your
network, and need to contact you network administrator.
5.7 Find Available Group & User ID Numbers
You need to find an available user ID number and an available group ID number.
These numbers must be available on both nodes. Make sure you have these
before continuing. The recommended user ID number is 201; the group ID
number is 210. To check these, simply enter smit user or smit group at the
command prompt, and enter List All at the menu. When you have done so, enter
this information on your worksheet.
5.8 Find the Correct Terminal Emulation
ORACLE_TERM is the Oracle terminal emulation variable. When you have found one,
enter it on your worksheet. The Oracle installer utility supports the following
terminal types:
3151, 386*, Q303*, Q310*, ansi, dec, hft*, hp, iris, lft, sun, vt100, vt220, wy50,
wy150, xlft, xsun, xsun50
* Multiple versions of these emulations
Once you have picked an emulation you think will work, you must test it. At the
command prompt, enter TERM= to whatever emulation you picked. Then export
TERM.
# TERM=lft
# export TERM
If everything seems to work fine, and you have the ability to backspace and
destroy old text, you will be OK with that terminal emulation for the installation.
Quick Install Program Preparation
39
5.6.1 Update .rhosts on Both Nodes
As the root user, edit the .rhosts file adding the names of all adapters involved in
cluster communication as well as the level of user allowed
Figure 7. .rhosts File from node1
5.6.2 Update /etc/hosts on Both Nodes.
As the root user, edit the /etc/hosts file adding the names and IP addresses of all
adapters involved in the cluster.
Figure 8. /etc/hosts file from node1 Preinstallation
38
ORACLE Cluster POWERsolution Guide
After completing the calculations on these sizing sheets, we must fill in the A, B,
and E totals from the individual sheets above into the final calculation sheet
shown below.
Table 17. Total Disk Sizing Result Sheet
Total Space Requirements (MB)
Transfer Table
Oracle Server Products
Precompiler Products:
Development Cycle
Precompiler Products:
User Cycle
Oracle Networking
Products
Oracle Tools Products
Oracle Online
Documentation
Additional installation
space for files
generated during
installation
Temporary space for
relinking
TOTALS
Disk Space
(A)
101.84
Database Space
(B)
12.5
Virtual Memory
(E)
15.68
n/a
n/a
35.71
0.2
1.36
7.22
0.6
n/a
4.25
n/a
+ 21.00
n/a
n/a
40.00
n/a
n/a
205.77
13.3
21.29
Note: From the above Total Disk Space Sizing Summary sheet, we can estimate
that the minimum space for those selected products for nonparallel database
installation on the first sample cluster is about 250 MB.
5.3 Copying Files From Floppy Disk
You must (as root user) copy the install program scripts from the included floppy
disk to the /tmp/hascripts directory.
# tar -xvf/dev/rfd0
5.4 Installation Worksheet
Make a copy of the worksheet from Appendix A that is appropriate for your
failover type. Fill out the information on this worksheet as we go along, and keep
it on hand. Much of the install is dependent on information from this worksheet.
5.5 Assign Network Adapters
If your network is not already established, you need to assign and name all of the
network adapters you will be using in the cluster. Each adapter has a label and an
IP address (as seen on your worksheet). For more information on this step please
contact your network administrator. Enter this information on your worksheet.
5.6 Update /.rhosts and /etc/hosts
These files control communication between adapters on different nodes. These
allow for name resolution as well as communication by IP address. The IP labels
used in these examples are the defaults used by the Quick Install Program.
Quick Install Program Preparation
37
Table 16. Disk Sizing Result Sheet for Oracle Networking Products
Space Requirements for Oracle Networking Products
Disk Storage Requirements
Memory Space Requirements
v
Dist.
DB Sp #1 User
Additional Users
Use
Product
(MB)
(MB)
(KB)
Users
KB per
Total
SQL*Net V1:
SQL*Net Async
0.27
N/A
SQL*Net DECnet
0.12
N/A
SQL*Net LU6.2
0.13
N/A
SQL*Net SPX/IPX
2.55
N/A
SQL*Net Named
N/A
Pipes
SQL*Net TCP/IP
0.32
N/A
SQL*Net V2:
v
SQL*Net V2
35.59
0.2
TNS Listener
1183
1
x 176
176
DECnet
0.12
N/A
LU6.2
0.14
N/A
7
x2
OSI
0.16
N/A
SPX/IPX
4.51
N/A
695
x 152
Names Server
3.93
N/A
1806
x 215
v
TCP/IP
0.12
N/A
Interchange
5.81
N/A
Network Manager
22.51
N/A
3266
x 323
Interchange Control Utility
1260
x 169
Interchange Listener
1224
x 168
Navigator
1187
x 168
Configuration Tool
N/A
0.2
ΣA
ΣB
ΣC
ΣD
Subtotals =
35.71
0.20
1183
176
Summary
A
Tot. Distribution Space (A)
35.71
(MB)
B
Tot. Database Space (B)
0.20
(MB)
C
#1 User Memory (C)
1183
(KB)
D
Additional User Memory (D)
176
(KB)
E
Virtual Memory Total (C+D)
1359
(KB)
Note: Each product selected for installation is indicated by
space required for each product is written in bold type.
36
ORACLE Cluster POWERsolution Guide
(v). The total disk
Table 15. Disk Sizing Result Sheet or Oracle Tools Products
v
Use
v
v
v
A
B
C
D
E
Space Requirements for Oracle Tools
Disk Storage Requirements
Memory Space Requirements
Dist.
DB Sp #1 User
Additional Users
Product
(MB)
(MB)
(KB)
Users
KB per
Total
Easy*SQL
Oracle Data Query
17.81
0.4
3424
x 756
Oracle*Mail
48.50
N/A
2628
x 362
Oracle*Terminal
5.77
N/A
667
x 194
PL/SQL 1.0
N/A
PL/SQL 2.0
0.28
N/A
SQL*Calc
2.99
N/A
Oracle Toolkit I
1.29
N/A
Oracle XA Library
0.15
N/A
SQL*Forms V3.0
24.99
1.2
Design (Char Mode)
4614
x 1034
Runtime (Char Mode)
3511
x 710
SQL*Menu V5.0
21.84
0.8
Design (Char Mode)
4388
x 998
Runtime (Char Mode)
3565
x 760
SQL*Plus
5.65
0.6
3472
1
x 778
778
SQL*QMX
2.97
SQL*Report
2.37
N/A
rpt
1748
x 230
rpf
46
x9
SQL*ReportWriter
2.37
N/A
Design
2642
x 230
Runtime
2424
x 302
ΣA
ΣB
ΣC
ΣD
Subtotals =
7.22
0.6
3472
778
Summary
Tot. Distribution Space (A)
7.22
(MB)
Tot. Database Space (B)
0.60
(MB)
#1 User Memory (C)
3472
(KB)
Additional User Memory (D)
778
(KB)
Virtual Memory Total (C+D)
4250
(KB)
Note: Each product selected for installation is indicated by
space required for each product is written in bold type.
(v). The total disk
Quick Install Program Preparation
35
Table 14. Disk Sizing Result Sheet for Oracle Server Products
v
Use
v
v
v
v
A
B
C
D
E
Space Requirements for Oracle Server Products
Disk Storage Requirements
Memory Space Requirements
Dist.
DB Sp #1 User
Additional Users
Product
(MB)
(MB)
(KB)
Users
KB per
Total
Oracle Server
36.60
12.5
6458
1
x 489
489
SQL*DBA
2589
1
x 283
283
SQL*Loader
1778
1
x 204
204
Export
1654
1
x 223
223
Import
1590
1
x 207
207
Server Manager
20.19
N/A
1752
x 221
(Line Mode only)
42.88
N/A
3515
x 364
Server Manager
(Motif bitmapped &
Line Mode)
+ Distributed Opt
0.10
N/A
+ Parallel Server
0.08
N/A
Opt.
+ Parallel Query
0.08
N/A
Opt
Toolkit II
43.06
N/A
Oracle Common
22.10
N/A
Lib & Utilities
Migration Utility
1.85
N/A
Oracle Parallel
30.00
10.0
1024
Backup/Restore
Util.
ΣA
ΣB
ΣC
ΣD
Subtotals =
101.84
12.5
14069
1612
Summary
Tot. Distribution Space (A)
101.84
(MB)
Tot. Database Space (B)
12.50
(MB)
#1 User Memory (C)
14069
(KB)
Additional User Memory (D)
1612
(KB)
Virtual Memory Total (C+D)
15681
(KB)
Note: Each product selected for installation is indicated by
space required for each product is written in bold type.
34
ORACLE Cluster POWERsolution Guide
(v). The total disk
Chapter 5. Quick Install Program Preparation
This chapter will get the cluster and user prepared to run the Quick Install
Program.
5.1 Choosing Oracle Products to Install
The Oracle products selected for our test installation on the Rotating, Hot
Standby, and Mutual Takeover configurations are listed below.
1. Oracle Server Products:
• Oracle7 Server with the Parallel Query option and its dependent products,
such as the Oracle Common Libraries and Utilities and Toolkit II.
2. Oracle Tools Products:
• SQL*Plus Release 3.3
3. Oracle Networking Products:
• SQL*Net V2.3
• TCP/IP Protocol Adapter V2.3
We did not install any CASE products, precompiler products, or Oracle on-line
documentation.
5.2 Disk Space Sizing for Selected Oracle Products
After selecting the products listed above (or any you chose), we now must
calculate the disk space requirement based on disk space sizing worksheets that
we described in Chapter 3, “Database Planning, System Requirements, and
Sizing” on page 15.
The install program accounts for all space requirements. In most cases, the sizes
used are double the minimum recommended. To make any adjustments, follow
the disk space sizing worksheets below:
© Copyright IBM Corp. 1997
33
4.3 Mutual Takeover Configuration
This cluster, consisting of two nodes, is set up in what is traditionally called a
Mutual Takeover configuration. The cluster is configured as shown in Figure 6.
user
Network:ETHER1
node1: 10.0.1.1
node1_boot: 10.0.1.11
node1_standby: 10.0.2.1
node2: 10.0.1.2
node2_boot: 10.0.1.12
node2_standby: 10.0.2.2
node2vg
node1vg
Nodename: node1
Network: TTY1
Nodename: node2
Figure 6. Cluster Configuration for mutual takeover Resources
Mutual Takeover is a form of cascading configuration where both nodes have their
own resource groups and application servers. There are two Oracle 7 databases
in the cluster, one contained in volume group node1vg and the other in volume
group node2vg. Node1 is the high-priority node for the database in node1vg, and
node2 is the backup. For the database in node2vg, node2 is the high-priority
node, and node1 is the backup. In the event that either node fails, the opposite
node acquires its resources. When the failed node reintegrates, its resources are
returned to it.
32
ORACLE Cluster POWERsolution Guide
4.2 Hot Standby Configuration
This cluster, consisting of two nodes, is set up in a Hot Standby configuration.
The cluster is configured as shown in Figure 5.
user1
service address
node1: 10.0.1.1
Network:ETHER1
node1_boot: 10.0.1.11
node2_boot: 10.0.1.12
node1_standby: 10.0.2.1
node2_standby: 10.0.2.2
node1vg
Node:node1
Network: TTY1
Node:node2
Figure 5. Cluster Configuration for Hot Standby Resources
4.2.1 Hot Standby Cluster Description
The Hot Standby cluster has a single resource group and application server. The
node1vg volume group contains a single Oracle 7 database. By default, node 1 is
assigned the high priority (server) role, and node2 is assigned the low priority
(backup) role. Whenever node1 joins the cluster, it acquires the resources and
starts the Oracle database. If node1 fails, the database is taken over by node2.
When node1 rejoins the cluster, it reacquires the resources from node2.
HACMP Cluster Configurations
31
4.1 Rotating Standby Configuration
This cluster configuration, consisting of two nodes, is set up in a rotating standby
configuration. The cluster was configured as shown in Figure 4.
user1
shared (rotating address)
node1: 10.0.1.1
Network:ETHER1
node1_boot: 10.0.1.11
node2_boot:10.0.1.12
node1_standby: 10.0.2.1
node2_standby: 10.0.2.2
node1vg
Network: TTY1
Node:node1
Node:node2
Figure 4. Cluster Configuration for Rotating Resources
The Rotating Standby cluster has a single resource group and application server.
A "shared" adapter (we call it shared, although in the actual HACMP configuration
panels it is called a service adapter) is configured for both nodes. The resource
group containing the Oracle database is acquired and started by the first node to
enter the cluster (start HACMP). On failure of the node that is serving the
resources, the other node acquires the resources and will not release them
unless it fails or leaves the cluster using the "graceful with takeover" option. If a
node leaves the cluster (gracefully or because of failure) and then rejoins, it
assumes the backup role until the other node leaves the cluster.
30
ORACLE Cluster POWERsolution Guide
Chapter 4. HACMP Cluster Configurations
This chapter describes the HACMP with Oracle 7 configurations that are available
using the Quick Install Program included with this redbook. The levels of software
that have been tested with the program are:
• AIX 4.2 and 4.2.1
• HACMP 4.2 and 4.2.1
• Oracle Server 7.3.2
In each case, the cluster consists of two nodes. The cluster configurations
available are fhe following:
• Rotating Standby
• Hot Standby
• Mutual Takeover
In each case, the configuration pictured uses the default settings in the Quick
Install Program for network addresses and labels, shared volume groups and so
on. These items are modifiable by the user.
© Copyright IBM Corp. 1997
29
28
ORACLE Cluster POWERsolution Guide
Space (A), in the final calculation sheet of disk space sizing before you prepare
your installation tasks.
For more information about this documentation library and how to use it, please
refer to the Getting Started with Oracle Online Documentation for AIX manual
from the Oracle Corporation.
3.3.3 Final Calculation of Disk Requirement
Enter the A, B, and E totals from the individual tables into the table below. Total
the three columns (including the additional Disk Space requirement) to determine
the total distribution space, database space, and virtual memory space you will
need.
Table 13. Total Disk Sizing Summary Sheet
Total Space Requirements (MB)
Transfer Table
Oracle Server Products
Precompiler Products:
Development Cycle
Precompiler Products:
User Cycle
Oracle Networking
Products
Oracle CASE Products
Oracle Tools Products
Oracle Online
Documentation
Additional installation
space for files generated
during installation
Temporary space for
relinking
TOTALS
Disk Space
(A)
Database Space
(B)
n/a
n/a
+ 21 MB
(Minimum)
Virtual Memory
(E)
n/a
n/a
n/a
n/a
+ 40 MB
(minimum)
Database Planning, System Requirements, and Sizing
27
text + data = 1st_User_Total
3. Multiply the data value from Step 1 by the number of additional users of the
program to determine the memory requirements for the remaining users of the
program (G).
(data) x (Total_Users -1) = Addtl_User_Total
4. Multiply total number of users of the program by the additional user memory
space requirement given for the Oracle server (from the Oracle and
Networking Products Sheets) to determine total user server space (H).
(Total_Users) x (Oracle_Server) = User_Server_Space
5. Add the result from steps 2-4 for the total space requirement for the program.
Enter the values for each program in the Totals column of the table below, and
add the column to produce the Total Memory Requirement (E).
Below are the tables showing the second-phase disk space sizing sheets for the
Oracle precompiler products.
Table 12. Second-Phase Disk Sizing Sheet for Oracle Precompiler Products
MEMORY SPACE REQUIREMENT
Precompiler Products: Use Cycle Calculation
#1 User
Addtl. Users
Executable Program
Text + Data
Data x (users-1)
Users x Server
(F)
(G)
(H)
Totals
Total Memory Req (E)
3.3.2.5 Oracle Product Installation Documentation Library Platform
The Product Installation Documentation Library contains port-specific, online
documentation for Oracle installation on your platform. It includes three kinds of
documentation: the Oracle7 Server for UNIX Administrator’s Guide, the Oracle
Tools for UNIX Administrator’s Guide, and the Installation & Configuration Guide
for your platform.
If you want to install this online product installation documentation library, you
must add about 18 MB to your Distribution Space (A) in the final calculation sheet
of the disk space sizing sheet before you prepare your installation tasks.
For more information about this documentation library and how to use it, please
refer to the Getting Started with Oracle Online Documentation for AIX manual
from the Oracle Corporation.
3.3.2.6 Oracle Product Documentation Library
The Product Documentation Library contains online documentation for each
Oracle product.
If you want to install this online product installation documentation library, you
must add from about 1.5 MB to a maximum of 40 MB of additional space,
depending on how many of the documents you want to install to your Distribution
26
ORACLE Cluster POWERsolution Guide
1. Calculate the Distribution (Dist) Space by subtotaling the values for the
selected products in A at the bottom of the column.
2. Calculate the Database space (DB Sp) by subtotaling the values for the
selected products in B at the bottom of the column.
3. Transfer the First User Memory Space (#1 User) value for the Oracle Tools to
the Subtotal C at the bottom of the column.
4. Calculate the Additional Users Memory Space.
5. Enter the number of additional users you estimate for the Oracle Tools in the
Users column.
6. Multiply Users by KB per and place the result in Total. For example:
7. 2 addtl. users of Pro*C V2 x 329 KB per = 658 KB
8. Transfer this value into Subtotal D at the bottom of the column.
9. In the table’s Summary section, enter A, B, C, and D subtotals.
10.Add the #1 User Memory (C) and Additional Users Memory (D), and enter the
result in the Virtual Memory Total (E).
Below are the tables showing the first-phase disk space sizing sheets for the
Oracle precompiler products.
Table 11. First-Phase Disk Sizing Sheet for Oracle Precompiler Products
v
Use
A
B
C
D
E
Space Requirements for CASE Products
Disk Storage Requirements
Mem3
ory Space Requirements
Dist.
DB Sp #1 User
Additional Users
Product
(MB)
(MB)
(KB)
Users
KB per
Total
Pro*Ada
1.10
N/A
3251
x 307
Pro*C v1.6
1.10
N/A
3261
x 396
Pro*C v2.2
1.10
N/A
3534
x 329
Pro*COBOL
1.10
N/A
3271
x 308
Pro*FORTRAN
1.10
N/A
3254
x 306
SQL*Module for Ada
8.83
N/A
3286
x 306
SQL*Module for C
8.83
N/A
3286
x 306
ΣA
ΣB
ΣC
ΣD
Subtotals =
Summary
Tot. Distribution Space (A)
(MB)
Tot. Database Space (B)
(MB)
#1 User Memory (C)
(KB)
Additional User Memory (D)
(KB)
Virtual Memory Total (C+D)
(KB)
After you have finished the first-phase calculation above, you must continue to
calculate the second phase, which is the use cycle calculation. To calculate the
minimum memory requirement for a finished user program, perform the following
steps:
1. Determine the text and data size of each executable by using the size
command.
2. Add the text and data total to determine the first user memory requirement for
the program (F).
Database Planning, System Requirements, and Sizing
25
Below are the tables showing the disk space sizing sheets for the Oracle CASE
products.
Table 10. Disk Sizing Sheet for Oracle CASE Products
v
Use
A
B
C
D
E
Space Requirements for CASE Products
Disk Storage Requirements
Memory Space Requirements
Dist.
DB Sp #1 User
Additional Users
Product
(MB)
(MB)
(KB)
Users
KB per
Total
CASE*Dictionary
53.18
7.5
Runtime
1730
x 237
CASE*Generator
33.65
4.5
Runtime
1895
x 281
CASE*Designer
15.11
3.5
casefront Runtime
2107
x 330
caseerd Runtime
2575
x 426
casefhd Runtime
2504
x 430
casedfd Runtime
2689
x 496
casemd Runtime
2904
x 930
ΣA
ΣB
ΣC
ΣD
Subtotals =
Summary
Tot. Distribution Space (A)
(MB)
Tot. Database Space (B)
(MB)
#1 User Memory (C)
(KB)
Additional User Memory (D)
(KB)
Virtual Memory Total (C+D)
(KB)
3.3.2.4 Oracle Precompiler Products
Although in this redbook we did not install any precompiler products, it is
necessary for you to know about disk size spacing for these products.
Calculating memory usage of these products also requires differentiation of
application development and production cycles.
During application development, a few programmers will be using the prox
(procob for COBOL, proc for C, proada for Ada, propas for Pascal, and profor for
FORTRAN) executable extensively. During the production cycles, many more
users may be executing the finished programs, which have been linked with the
SQLLIB run-time library. Remember that each precompiler application may
spawn an Oracle shadow process; so the same rule for including the
per-additional-user memory consumption of the Oracle Kernel applies to
precompilers.
Disk space for the precompilers is somewhat redundant. It is calculated assuming
only one language is installed on your system. When a second language is
installed, the additional free space required is not as great as listed because
different languages share common libraries.
The calculation procedures for sizing space requirements of Oracle precompiler
products have two phases. First, you have to calculate the development cycle
calculations, which are the same as for other products above.
1. Select the product you want to install by placing a check mark in the first
column of each product.
24
ORACLE Cluster POWERsolution Guide
casefhd
The diagrammer invoked by all users at the function hierarchy
definition stage of development.
casedfd
The diagrammer invoked by all users at the dataflow definition stage of
development.
casemd
The diagrammer invoked by all users who are checking completeness
and validity links between pairs of element types.
The calculation procedures for sizing space requirements of Oracle CASE
products are the same as above. The numbers given in this sheet are to be
understood as the bare minimum.
1. Select the product you want to install by placing a check mark in the first
column of each product.
2. Calculate the Distribution (Dist) Space by subtotaling the values for the
selected products in A at the bottom of the column.
3. Calculate the Database space (DB Sp) by subtotaling the values for the
selected products in B at the bottom of the column.
4. Transfer the First User Memory Space (#1 User) value for the Oracle Tools to
the Subtotal C at the bottom of the column.
5. Calculate the Additional Users Memory Space.
Enter the number of additional users you estimate for the Oracle Tools in the
Users column.
Multiply Users by KB per and place the result in Total. For example:
3 addtl. users of CASE*Designer casemd x 930 KB per = 2790 KB
Transfer this value into Subtotal D at the bottom of the column.
6. In the table’s Summary section, enter A, B, C, and D subtotals.
Add the #1 User Memory (C) and Additional Users Memory (D), and enter the
result in the Virtual Memory Total (E).
Database Planning, System Requirements, and Sizing
23
Below are the tables showing the disk space sizing sheets for the Oracle Tools
products.
Table 9. Disk Sizing Sheet for Oracle Tools Products
v
Use
A
B
C
D
E
Space Requirements for Oracle Tools
Disk Storage Requirements
Memory Space Requirements
Dist.
DB Sp #1 User
Additional Users
Product
(MB)
(MB)
(KB)
Users
KB per
Total
Easy*SQL
Oracle Data Query
17.81
0.4
3424
x 756
Oracle*Mail
48.50
N/A
2628
x 362
Oracle*Terminal
5.77
N/A
667
x 194
PL/SQL 1.0
N/A
PL/SQL 2.0
0.28
N/A
SQL*Calc
2.99
N/A
Oracle Toolkit I
1.29
N/A
Oracle XA Library
0.15
N/A
SQL*Forms V3.0
24.99
1.2
Design (Char Mode)
4614
x 1034
Runtime (Char Mode)
3511
x 710
SQL*Menu V5.0
21.84
0.8
Design (Char Mode)
4388
x 998
Runtime (Char Mode)
3565
x 760
SQL*Plus
5.65
0.6
3472
x 778
SQL*QMX
2.97
SQL*Report
2.37
N/A
rpt
1748
x 230
rpf
46
x9
SQL*ReportWriter
2.37
N/A
Design
2642
x 230
Runtime
2424
x 302
ΣA
ΣB
ΣC
ΣD
Subtotals =
Summary
Tot. Distribution Space (A)
(Mb)
Tot. Database Space (B)
(Mb)
#1 User Memory (C)
(Kb)
Additional User Memory (D)
(Kb)
Virtual Memory Total (C+D)
(Kb)
3.3.2.3 Oracle CASE Products
Although in this redbook we did not install any CASE products, it is necessary for
you to know about disk size spacing for these products.
The categories of space requirements given for CASE*Designer are as follows:
casefront The executable run by all users at all times. It is the startup window
from which users invoke the different CASE*Designer diagrammers,
depending on the stage of development.
caseerd
22
The diagrammer invoked by all users at the entity relationship
definition stage of development.
ORACLE Cluster POWERsolution Guide
v
v
A
B
C
D
E
Space Requirements for Oracle
SQL*Net V2
11.08
0.2
TNS Listener
DECnet
0.13
N/A
LU6.2
0.18
N/A
OSI
0.16
N/A
SPX/IPX
1.77
N/A
Names Server
6.59
N/A
TCP/IP
0.15
N/A
Interchange
7.30
N/A
Interchange Control Utility
Interchange Listener
Navigator
Data Pump
N/A
N/A
Configuration Tool
N/A
0.2
ΣA
ΣB
Subtotals =
Summary
Tot. Distribution Space (A)
Tot. Database Space (B)
#1 User Memory (C)
Additional User Memory (D)
Virtual Memory Total (C+D)
Networking Products
1710
7
695
3221
1302
2127
926
1001
x 94
x2
x 152
x 340
x
x
x
x
70
172
56
56
ΣC
ΣD
(MB)
(MB)
(KB)
(KB)
(KB)
3.3.2.2 Oracle Tools
The calculation procedures for sizing the space requirements of Oracle Tools
products is the same as above. You must keep in mind that First User Memory
(#1 User) is extremely difficult to predict because it depends heavily on the
application. Furthermore, the number of cursors opened by applications can
significantly affect the size of associated shadow processes. The numbers given
in this sheet are to be understood as the bare minimum.
1. Select the product you want to install by placing a check mark in the first
column of each product.
2. Calculate the Distribution (Dist) Space by subtotaling the values for the
selected products in A at the bottom of the column.
3. Calculate the Database Space (DB Sp) by subtotaling the values for the
selected products in B at the bottom of the column.
4. Transfer the First User Memory Space (#1 User) value for the Oracle Tools to
the Subtotal C at the bottom of the column.
5. Calculate the Additional Users Memory Space.
6. Enter the number of additional users you estimate for the Oracle Tools in the
Users column.
7. Multiply Users by KB per and place the result in Total. For example:
8. 9 additional users of SQL*Plus x 778 KB per = 7002 KB
9. Transfer this value into Subtotal D at the bottom of the column.
10.In the table’s Summary section, enter A, B, C, and D subtotals.
11.Add the #1 User Memory (C) and Additional Users Memory (D), and enter the
result in the Virtual Memory Total (E).
Database Planning, System Requirements, and Sizing
21
Enter the number of additional users you estimate for the Oracle Server in the
Users column.
Multiply Users by KB per and place the result in Total. For example:
9 users x 186 KB per = 1674 KB
Transfer this value into Subtotal D at the bottom of the column.
6. In the table’s Summary section, enter A, B, C, and D subtotals.
Add the #1 User Memory (C) and Additional User Memory (D), and enter the
result in the Virtual Memory Total (E).
Below are the tables showing the disk space sizing sheets for the Oracle server
products and Oracle networking products.
Table 7. Disk Sizing Sheet for Oracle Server Products
Space Requirements for Oracle Server Products
Disk Storage Requirements
Memory Space Requirements
v
Dist.
DB Sp #1 User
Additional Users
Use
Product
(MB)
(MB)
(KB)
Users
KB per
Total
v
Oracle7 Server
32.37
12.5
6978
x 521
SQL*DBA
2589
x 283
SQL*Loader
1873
x 102
Export
1741
x 121
Import
1675
x 105
Server Manager
12.57
N/A
1908
x 120
(Line Mode only)
Server Manager
12.57
N/A
5058
x 486
(Motif bitmapped &
Line Mode)
+ Distributed Opt
0.10
N/A
+ Parallel Server
0.08
N/A
Opt.
v
+ Parallel Query Opt
0.08
N/A
v
Toolkit II
68.39
N/A
v
Oracle Common Lib
31.48
N/A
& Utilities
Migration Utility
3.00
N/A 1816
x 247
Oracle Parallel
30.00
10.0
1024
Backup/Restore Util.
ΣA
ΣB
ΣC
ΣD
Subtotals =
Summary
A Tot. Distribution Space (A)
(MB)
B Tot. Database Space (B)
(MB)
C #1 User Memory (C)
(KB)
D Additional User Memory (D)
(KB)
E
Virtual Memory Total (C+D)
(KB)
Table 8. Disk Sizing Sheet for Oracle Networking Products
Space Requirements for Oracle Networking Products
Disk Storage Requirements
Memory Space Requirements
v
Dist.
DB Sp #1 User
Additional Users
Use
Product
(MB)
(MB)
(KB)
Users
KB per
Total
SQL*Net V2:
20
ORACLE Cluster POWERsolution Guide
3.3 Disk Space and Memory Sizing for Oracle
This section is designed as a planning tool. It provides the information required to
calculate the distribution space, database space, and memory requirements for
each Oracle product available for installation on your system. Also we will discuss
the memory requirement based on users. Follow the step-by-step procedure to
calculate the disk storage and memory requirements for your specific installation.
Note that these instructions provide only generalized estimates of your disk and
memory needs, not a precise calculation.
3.3.1 Required Calculations
The calculations result in two values for total disk storage and two values for
system memory. These values are:
Distribution Space
Space required in megabytes to store the products
on disk.
Database Space
Megabytes of space the product requires in the
database space. Assess the number of Oracle
blocks required by dividing this amount by the
default Oracle block size for your machine.
First User Memory
An estimate of virtual memory space, in kilobytes,
used when the first user runs the product.
Memory/Additional User
An estimate of the virtual memory space, in
kilobytes, used when each additional user runs the
product.
3.3.2 Disk Space Requirements
In this section, we will explain the disk space sizing sheet for use as a tool for
sizing and estimating the space required for an Oracle installation. There are six
kinds of sheets, one for each category of Oracle product.
3.3.2.1 Oracle Server and Networking Products
To arrive at a basic estimate of your space needs for Oracle and the networking
products, perform these calculations:
1. Select the product you want to install by placing a check mark in the first
column of each product.
Note: For server installation, several products, such as Oracle Server and its
components, Oracle Common Libraries and Utilities, Oracle Toolkit, and
SQL*Net, are mandatory.
2. Calculate the Distribution (Dist) Space by subtotaling the values for the
selected products in “A” at the bottom of the column.
3. Calculate the Database Space (DB Sp) by subtotaling the values for the
selected products in “B” at the bottom of the column.
4. Transfer the First User Memory Space (#1 User) value for the Oracle Server to
the Subtotal “C” at the bottom of the column.
5. Calculate the Additional Users memory space under the Memory Space
Requirements heading.
Database Planning, System Requirements, and Sizing
19
• Precompilers
• CASE Products
• SQL*Net V2
Note that when you install an individual product, the dependent products are
installed automatically by the Installer.
You can find more information about prerequisites or requirements for Oracle
products that are included in one of the four categories mentioned above in
Chapter 2, “Requirements,” in the Oracle7 for AIX-Based Systems Installation
and Configuration Guide.
Below, we only show you some examples of the tables that we installed in this
book, such as the Oracle Tools, PL/SQL Release 2.3 and SQL*Plus Release 3.3,
and the SQL*Net with TCP/IP Protocol Adapter requirements or prerequisites for
installation.
PL/SQL Release 2.3
Table 3. Prerequisites for PL/SQL Release 2.3
Software Requirement
Oracle Server
Hardware Requirement
None
Version
7.3
State During Installation
Installed
SQL*Plus Release 3.3
Table 4. Prerequisites for SQL*Plus Release 3.3
Software Requirement
Oracle Server
Hardware Requirement
None
Version
7.3
State During Installation
Running
SQL*Net Release 2.3
Table 5. Prerequisites for SQL*Net Release 2.3
Software Requirement
Version
State During Installation
Oracle Server
7.3
Installed
Oracle TCP/IP Adapter
2.3
Installed
TCP/IP for AIX 4.1.4
Included in O/S
Installed
Hardware Requirement
Token-Ring Adapter or Ethernet Adapter supported for IBM RS/6000
Oracle TCP/IP Adapter Release 2.3
Table 6. Prerequisites for Oracle TCP/IP Adapter Release 2.3
Software Requirement
Version
State During Installation
Oracle Server
7.3
Installed
SQL*Net
2.3
Installed
Hardware Requirement
Token-Ring Adapter or Ethernet Adapter supported for IBM RS/600
18
ORACLE Cluster POWERsolution Guide
3.1.4 File Locations
Several files that will be prompted for when you install and create your database
or object database must be put in the external shared disk if you need a highly
available database within an HACMP environment. The files to be put into the
external shared disk are as follows:
• Control file
• Redo log
• System tablespace
• Rollback segment
• Tools tablespace
• User tablespace
• Temporary tablespace
3.2 System Requirements
This section lists the hardware and software requirements for installing your
Oracle database.
3.2.1 Requirements for Oracle Product Installation on a Single Server
This section outlines the requirements for installing various Oracle products.
• Hardware Requirements
Table 1. Hardware Requirements for Oracle Product Installation on a Single Server
Hardware Item
Requirement
CPU
Memory
IBM RISC System/6000.
Please refer to the disk space and
memory requirement chart in the next
section.
An 8 mm tape drive is required for the
tape distribution of Oracle. The tape
drive should be set to 512 block size.
ISO 9660 format is supported.
The following terminal types are
supported: ANSI, HFT, LFT, 3151,
VT100, WYSE50, and so on.
SQL*Net TCP/IP requires an adapter
that will support TCP/IP.
Tape Device
CD-ROM
Terminal
Controller
• Software Requirements
Table 2. Software Requirements for Oracle Product Installation on a Single Server
Software Item
Requirement
Operating System
AIX Version 4.1.4 or later
3.2.2 Requirements for Oracle Optional Products
For the prerequisite requirements of Oracle optional products, please refer to the
Oracle manual. The Oracle products are usually grouped into the following
categories:
• Tools
Database Planning, System Requirements, and Sizing
17
3.1.2.1 OLTP Database
An OLTP database, such as a bank automated teller machine or ATM, has a very
high volume of transactions (measured in transactions per second). When
planning a database for such purposes, you usually need to:
• Split database files so that disk I/O is shared between many disk subsystems
• Split the logical database design
• Implement redo log archiving
• Be available 7 days a week, 24 hours a day
• Use raw disk partitions
• Use high availability hardware (mirrored disks, for example)
• Institute performance tuning
• Set initsid.ora parameters larger than the default values to accommodate
more users and increase performance
3.1.2.2 Decision Support Database
A decision support database, such as an inventory system, has a relatively low
number of database updates (measured in transactions per hour). The users tend
to make few queries, and they look at the results of these queries for many
minutes at a time. They may prepare reports from these queries. A decision
support database:
• Has less need to distribute I/O across multiple disks and controllers
• Has less need to split the logical database design
• Has less need to implement redo log archiving
• Can usually be brought down for maintenance or backups
3.1.2.3 Development Database
A development database is used to develop new applications. Such a database:
• Does not usually need to split I/O
• Can usually be brought down for maintenance or backups
• Has little need for backup more than once a day
• Has little need for performance tuning
3.1.2.4 Mixed Database
A mixed database combines various functions, such as decision support and
transaction processing. Most databases fall within this category. Such a
database:
• Has an even mixture of queries and updates
• Has some performance tuning needs, but not as much as an OLTP database
3.1.3 Database Size
When considering the size of a database, you must always consider it relative to
the system upon which it will run. The basic considerations are:
• A machine with more physical memory can support a larger database
• A machine with more processors can support a larger database
16
ORACLE Cluster POWERsolution Guide
Chapter 3. Database Planning, System Requirements, and Sizing
This chapter provides information for planning different types of databases. It also
discusses AIX-specific issues, including system hardware and software
requirements. Included is disk sizing information for your database.
3.1 Planning Your Database
The following section provides information for planning different types of
databases.
3.1.1 First Considerations
When planning database installations, you need to determine your processing
priorities and the amount of system resources you have.
3.1.1.1 Priorities
Your priorities will decide the characteristics of your database. Things to consider
are:
• Is transaction speed more important than data safety?
• Will the database be used for a large number of short transactions or for a
small number of large transactions?
• Is the database to be accessed by a single application or by multiple
applications running on multiple nodes?
3.1.1.2 System Resources
The system resources available to you will also affect your database installation,
usage and backup strategies. The following considerations are important:
• The number of users and applications running at one time will depend upon
the physical memory available and the system running the database.
• You need to know how many disk drives are available, their device names, the
speeds of the various disks (are some faster than others), and the number of
disk controllers available.
• For backup purposes, you should know the number and type of tape drives
available.
• What UNIX kernel parameters need to be changed or added?
• What network resources are available?
3.1.2 Database Application Type
The type of database application you are planning is a primary criterion in
determining how you will plan and structure the database. Depending on the
purpose for which they will be used, there are four basic types of database
applications:
• On Line Transaction Processing (OLTP)
• Decision Support
• Development
• Mixed
Database Planning, System Requirements, and Sizing
15
checkpoint, all of the database file headers and the redo log file headers are
updated to record the fact that a checkpoint has occurred.
The CKPT is optional. If it is not present, the LGWR process performs the tasks
of the CKPT process. It is recommended that the CKPT process is enabled when
there are a large number of data files.
2.3.6 Recover (RECO) Process
The recover process is used when there is a failure in a distributed transaction. A
distributed transaction is one where two or more locations of the data have to be
kept synchronized. In this environment, there may be multiple databases on
multiple interconnected servers and either a node or a network fails.
Any transaction that may have completed in one site but not in another is referred
to as in doubt. The RECO process attempts to establish communication with the
remote servers. When the connection is reestablished, the RECO process
automatically resolves all the in doubt transactions.
The RECO process is optional and is only needed in instances that carry out
distributed transactions.
2.3.7 Archiver (ARCH) Process
The archiver optional process is used when the data is running in ARCHIVELOG
mode and automatic archiving is enabled. ARCH copies the redo entries from the
online redo log files to the archive area.
2.3.8 Parallel Server Lock Processes (LCKn)
The lock processes are used only if running Oracle Parallel Server. LCKn uses
the distributed lock manager for inter-instance locking to prevent simultaneous
changes to the same data from different instances.
14
ORACLE Cluster POWERsolution Guide
There are four Oracle system background processes that must always be up and
running for the database to be usable.They are the following:
1. Database Writer (DBWR)
2. Log Writer (LGWR)
3. System Monitor (SMON)
4. Process Monitor (PMON)
2.3.1 Database Writer (DBWR)
The database writer process is responsible for writing modified (dirty) data blocks
from the database buffer cache to the database files. The DBWR only does a
write when buffers are needed for data to be read in. There may be more than
one DBWR process.
When using the Oracle Parallel Server, the DBWR might be forced to do a write,
not because buffers are needed but because another user needs to modify the
same buffers. This process is known as pinging and can be responsible for
degrading database performance.
2.3.2 Log Writer (LGWR)
The log writer process writes redo log entries from the redo log buffer to the redo
log files on disk.
2.3.3 System Monitor (SMON)
The system monitor process performs instance recovery at database startup time
or, in the case of Oracle parallel server, when another instance belonging to the
database has crashed or terminated abnormally.
SMON also releases temporary segments that are no longer needed, compacts
the free space fragments in the database files, and detects deadlock situations.
2.3.4 Process Monitor (PMON)
The process monitor process keeps track of database processes. If a user
process fails, PMON cleans up the cache and frees up any resources that the
failed process was using.
PMON also monitors the dispatcher and shared server processes and restarts
them if necessary.
Along with the four processes described above, there are a number of other
optional system processes, such as:
• Checkpoint (CKPT)
• Recover (RECO)
• Archiver (ARCH)
• Parallel Server Lock Processes (LCKn)
2.3.5 Checkpoint (CKPT)
A checkpoint is an event in which all modified data blocks are written by the
DBWR process to the data files. This usually occurs at a redo log switch. At a
Oracle Concepts
13
User
User
User
User
User
User
Listener
Dispatcher Dnnn
Shared server
processes
Dedicated server
processes
SYSTEM GLOBAL AREA (SGA)
Database
buffer cache
Redo log
buffer
PMON
Shared
pool
SMON
RECO
DBWR
CKPT
LGWR
LCKn
ARCH
Database
files
Figure 3. Oracle Architecture
12
ORACLE Cluster POWERsolution Guide
Redo log
files
Archive
log files
• Stack space: memory holding session variables
• Session information: stored here for dedicated servers and in the SGA for
mutlithreaded servers
• Private SQL area: information on binding variables and run-time buffers
2.3 Oracle Processes
An Oracle instance consists of:
• User processes
• Oracle server processes (shadow processes)
• Oracle background processes
• Shared memory used by these processes
The database can be organized further to run with a dedicated server or with a
multithreaded server. In the dedicated server configuration, each user process
has a dedicated server process. In the multithreaded server configuration,
multiple user processes are serviced by a few shared server processes. Figure 3
on page 12 provides a graphical depiction of the Oracle architecture.
Oracle Concepts
11
Whenever there is a change to the structure of the database, the control file is
updated.
2.1.4 The init.ora File
The init.ora file is a parameter file (text) containing Oracle system parameters. It
is provided by Oracle and should be customized for your site. This file is read
during database startup to determine the size of the system global area (see
Section 2.2.1, “System Global Area or Shared Global Area (SGA)” on page 10)
and to locate the control files. The actual name of the init.ora file has the Oracle
instance identifier appended to it.
2.2 Oracle Memory Structure
Oracle uses system memory to run user processes, cache data, and indexes, and
to store shared program code. Oracle has two types of memory structures:
3. The system global area
4. The program global area
2.2.1 System Global Area or Shared Global Area (SGA)
The SGA is an area of shared memory used by Oracle to store data and control
information for one Oracle instance. All of the information contained within the
SGA is shared by all of the users connected to the instance. The SGA is allocated
when the instance starts up and is automatically deallocated when the instance
shuts down.The SGA is made up of the following key components:
2.2.1.1 Data Buffer Cache
The data buffer cache stores the most recently used data blocks— that is, those
blocks which have been modified but not yet written to disk (dirty blocks) as well
as those which have been written to disk (clean blocks).
Before a user process can access a piece of data, it has to be in the data buffer
cache. A least recently used algorithm is used to free up space when new data is
requested by a user.
2.2.1.2 Redo Log Buffer
Before any transactions can be recorded into the redo logs, they must first reside
in the redo log buffer. They are then written to the redo logs by a database
background process (log writer).
2.2.1.3 Shared Pool
The shared pool is a cache containing all the parsed SQL statements that are
ready to run. This is useful for reducing overheads (memory, processing time,
execution planning time) when multiple applications issue the same SQL
statement.
2.2.2 Program Global Area (PGA)
A PGA is allocated when a user connects to the database. The PGA is not shared
and contains data and control information for single Oracle server processes or
for Oracle background processes. The PGA is made up of the following
components:
10
ORACLE Cluster POWERsolution Guide
Temporary Segment: This is used by Oracle when an SQL statement needs a
temporary work area. It is destroyed upon completion of the statement.
Data and Index Segments: Data segments store user data within the database.
Index segments store indexes used by Oracle to look up data quickly. An index
scan of the database is much quicker than a full table scan, where Oracle would
look at every row in the database.
2.1.2 Redo Log Files
The redo log files are journal files that record all changes made to the database.
These files are in memory (for performance reasons, disk I/O is roughly a
thousand times slower than actions in memory), and Oracle can then write the
changes to the data files (on disk) at its leisure. Every Oracle database has a set
of two or more log files.
These files are used in case a failure (machine, disk, and so forth) prevents
modified data from being permanently written to the data files. During recovery,
Oracle will apply the data in these files to bring the database to a consistent state
without losing any committed transactions.
2.1.2.1 Online Redo Logs
Since these files are open or online during normal operation, they are referred to
as the online redo log files. The online redo logs work in a circular fashion. As
transactions take place, they are recorded in the first redo log. When this is full, a
log switch occurs. Now, transactions are recorded in the second redo log. When
this is full, another log switch occurs. All new transactions are once again
recorded in the first log, overwriting the previous contents.
2.1.2.2 Archived Redo Logs
Oracle offers the possibility of running in either ARCHIVELOG mode or
NOARCHIVELOG mode.
In ARCHIVELOG mode, the contents of the redo logs are copied to an archive
area (on disk) before they are overwritten. These archive files are known as the
archived redo log files. Because they are not open during normal operation of the
database and are only required during data recovery, they are also known as the
offline redo logs.
In NOARCHIVELOG mode, old redo logs are not kept, and the redo logs are
simply overwritten.
2.1.3 Control File
Every Oracle database has a control file. It is highly recommended to have more
than one copy to guard against data loss. This file records the physical structure
of the database and contains the following types of information:
•
•
•
•
•
Name of the database
Creation date and time
Location of the database
Status and state of all the data files
Location of the redo logs
Oracle Concepts
9
DATABASE
Tablespace
Tablespace
Figure 2.
Datafile
Datafile
table
table
Logical Structure of a Typical Oracle Database
Database files have the following characteristics:
• A data file is only associated with one database.
• The size of a data file cannot be changed after its creation.
• One or more data files can be grouped together to form a tablespace.
2.1.1.1 Tablespaces
A database is made up of one or more tablespaces. Tablespaces are useful in
helping you to organize your data for ease of management, security and
performance.
System Tablespace is used to store the data dictionary. This holds information
such as names of tablespaces and what datafiles are contained in each one.
User Tablespace(s) are used to hold personal data.
2.1.1.2 Segments
A segment is a generic name given to any object that occupies storage in the
database. A segment is made up of a group of extents (contiguous blocks) that
are in the same tablespace (but not necessarily in the same data file).
Extents are used to minimize the amount of wasted storage and can grow or
shrink as required. Extents themselves are made up of data blocks, which are the
smallest pieces that make up an Oracle database and are physically related to
the disk partition size. An Oracle database has the following types of segments:
Rollback Segment: Whenever data is altered, this change must either be
committed or rolled back. The rollback segment holds the previous version of the
data being modified; this allows for recovery from aborted or incomplete
transactions.
8
ORACLE Cluster POWERsolution Guide
Chapter 2. Oracle Concepts
The Oracle database follows the three-tiered database structure outlined in
Chapter 1, “Relational Database Concepts” on page 1. An Oracle database is
generally divided into tablespaces and schema objects such as tables, views and
indexes. A description of the three levels as well as the relationship between
them is stored in the data dictionary.
This chapter describes the more important Oracle files that will be installed on
your system and also gives a outline of the architecture of an Oracle instance.
2.1 Oracle Files
An installation of Oracle on your system consists of the following sets of files:
• Data Files
• Redo Log Files (online and archive)
• Control Files
• Initialization Parameter File (INIT.ORA)
2.1.1 Data Files
An Oracle database contains one or more data files. The data files make up the
physical repository for all information stored in the database. Two main groups of
information are stored in the data files:
• User-created objects (tables, indexes, clusters)
• The data dictionary that is found in the SYSTEM data file created during the
Oracle installation process
Figure 2 shows the logical structure of a sample Oracle database.
© Copyright IBM Corp. 1997
7
6
ORACLE Cluster POWERsolution Guide
server is badly managed. There may also be problems arising from memory
being overwritten by nondatabase applications.
1.5 Filesystems and Raw Logical Volumes
The files of a database provide the actual physical storage for database
information. All of the databases used allow data to be stored either as files within
a filesystem or directly on the disk through the raw device interface in logical
volumes.
1.5.1 Filesystems
Using filesystem files is easier than using raw devices. In this case, the database
simply creates a file in the filesystem. However, when using the filesystem, the
database must contend with the disk caching that is inherent in block devices.
Block devices do buffered I/O, where data is collected in a buffer until an entire
block can be transferred at one time. Generally, writes need to be done
synchronously in order to ensure coherency.
On reads, there are overheads due to the data being read into the disk cache
before becoming exploitable by the database.There is also the problem that the
files in the filesystem may not be placed contiguously on the disk, and additional
overheads will be incurred due to disk seek times.
1.5.2 Raw Logical Volumes
A raw logical volume is an area of contiguous physical and logical disk space that
is under the direct control of an application rather than under control of the
operating system and filesystem.
The applications use character (raw) input and output, carrying out a data
transfer with every read or write rather than the block input and output of
filesystems. As data is written directly to the disk, bypassing all disk caching and
filesystem overheads, performance is generally improved.
Backup and recovery of raw devices is more complicated because file copies do
not work. Backups are carried out by using either the UNIX dd command or a
third-party backup utility.
Relational Database Concepts
5
1.4 Client/Server Architecture
In a database using client/server architecture, the database application is divided
into two parts:
1. A front-end or client portion
2. A back-end or server portion
The client part executes the database front-ends, such as interactive query tools
or report writers or the application that accesses the database. It often interacts
with the user through a graphical user interface. The clients make SQL requests
to the server.
The server part executes the database-management-system software and
handles all of the functions required for data access and storage. The server
processes the request sent to it by the client, calculates the result, and sends the
data back to the client.
Although both the client and server parts can execute on the same machine, it is
more efficient when client and server are separated in a networked environment.
1.4.1 Client/Server Connection
The client/server connection can be made in various ways. A connection here
means the logical association between two applications, such as a client
application and the database server. The connection must be established before
data transfer and must be maintained for the duration of the transfer.
Most databases use two types of connections:
1. Network connections
2. Shared memory connections
1.4.2 Network Connections
Network connections themselves may be of two types.
1.4.2.1 Remote Network Connections
Remote network connections are used generally when the client and server are
on different computers. In this case, the connection is made across a physical
network linking the two machines.
1.4.2.2 Local Loopback Network Connections
When the client and server are on the same machine, the setup is called a local
loopback connection. The connection is established using the networking
facilities of the machine and behaves as though the client and server are on two
separate machines.
1.4.3 Shared-Memory Connections
Shared-memory connections can only be established when the client and the
server are on the same machine. The client and server share the system memory
for very fast data transfer. This type of connection can be more prone to data
integrity problems if the shared-memory communication between client and
4
ORACLE Cluster POWERsolution Guide
Queries are statements that retrieve (but do not change) data and begin with the
reserved word SELECT.
Data Manipulation statements are used to change data by using instructions such
as:
• INSERT—to add new rows of data
• DELETE—to delete rows of data
• UPDATE—to change column values in existing rows
Data Definition statements are used to create, alter and drop database objects.
They differ from the previous two types of statements because they require a
write access to the data dictionary. Typical reserved words are CREATE TABLE,
DROP TABLE, and ALTER TABLE.
Access Control statements are used for two types of access:
• Access to the database system
• Access to the database data
Access control statements like GRANT and REVOKE can control user access
privileges.
Transaction Control statements are of two types:
• COMMIT—to end the current transaction
• ROLLBACK—to abort the current transaction
1.3.2 Context Areas and Cursors
Each SQL statement is associated with a cursor or context area. This is a
memory buffer created on the server that contains the current status of one SQL
statement.
A cursor can be thought of as a file handle (or name of the context area) that is
opened to gain access to the SQL statement results. The statement is processed
or parsed within the cursor, and as long as the cursor remains open, the
statement can be re-executed without being reparsed.
Just as a file handle keeps track of the program’s current position within a file, the
cursor keeps track of the state or phase of the statement. Cursors are only
capable of forward sequential processing. When the statement has been fully
parsed and no longer needs to be executed, the cursor is closed either explicitly
or by the client terminating the connection to the server.
1.3.3 Statement Parsing
An SQL statement is parsed (processed), and a representation of the statements
is loaded into the cursor. Parsing consists of:
•
•
•
•
•
•
Translating the SQL statement
Loading it into a cursor
Verifying user access privileges (access control)
Determining access paths to be used when executing the statement
Determining resource requirements
Reserving the resources required
Relational Database Concepts
3
Although the information stored in each table may be independent, there is
obviously a link between the two tables shown. The department ID in the
Employees table is “related” to the department ID in the Managers table. In order
to find the name of the manager of a particular employee, you look up his/her
department ID (in the Employees table) and then find the manager of this
department (in the Managers table).
Relational databases are data driven; that is, the tables and relationships
between them are defined once, and then only the data changes over time. Data
changes have no impact on the applications using them.
1.2 Database Structure
The structure of a relational database can be divided into three parts:
1. Physical
2. Logical
3. External
The physical part is made up of the files, the directories, and the physical storage
elements.The logical part is made up of the objects that are referenced by the
database. These include tables, tablespaces, and other elements that make up
the relational database model. The boundaries of logical structures are
independent of physical space allocation. The external part represents the data
as seen by the users. This includes views of the data, the clustering of the
information, and accessibility to the various tables.
This three-tiered approach allows for an independence between the data and the
means of storage of the data, and it allows for different types of access to the
data depending upon the needs of the user.
1.3 Structured Query Language (SQL)
SQL (pronounced sequel or S.Q.L.) is an “English-like,” non-procedural language
that is used for most database actions. It was developed by IBM research and has
since been refined by ANSI.
1.3.1 SQL Statements
The SQL language consists of about 30 statements. Each SQL statement begins
with a verb (which describes what the statement does) followed by one or more
clauses.
The main types of SQL statements are:
• Queries
• Data Manipulation
• Data Definition
• Access Control
• Transaction Control
2
ORACLE Cluster POWERsolution Guide
Chapter 1. Relational Database Concepts
A database is a means of organizing a collection of data. The basic tasks
required of a database are:
• Entering data
• Storing data
• Manipulating data
This chapter provides an overview of the features that are common to all of the
databases referred to in this redbook. The following topics are discussed:
• Relational databases
• Database structure
• Structured Query Language
• Client/Server architecture
• File systems and raw logical volume
1.1 Relational Databases
All of the database management systems discussed in this redbook are
relational. In this type of database, the information is stored in tables. Each table
has a name and is logically represented as being made up of rows and columns.
Tables can be related to each other if they contain columns with common types of
information.
.
Name
Dept. ID
John
Relationship
Dept. ID
Manager
0123
0123
Wally
Jane
0333
0135
Helen
Sally
0172
0172
Harry
Winston
0231
0231
Betty
0333
James
0450
Neil
TABLE 1: EMPLOYEES
TABLE 2: MANAGERS
Figure 1. Relationship between Two Tables
© Copyright IBM Corp. 1997
1
We want our redbooks to be as helpful as possible. Please send us your
comments about this or other redbooks in one of the following ways:
• Fax the evaluation form found in “ITSO Redbook Evaluation” on page 117 to
the fax number shown on the form.
• Use the electronic evaluation form found on the Redbooks Home Pages at the
following URLs:
For Internet users: http://www.redbooks.ibm.com
For IBM Intranet users: http://w3.itso.ibm.com/redbooks
• Send us a note at the following address:
[email protected]
xiv
ORACLE Cluster POWERsolution Guide
Preface
This redbook describes the implementation of the Oracle database products in a
highly available AIX cluster using HACMP. Its purpose is to provide
documentation for a Quick Install Program that automates much of the
procedures for configuring an HACMP cluster by creating necessary shared disk
areas and installing the Oracle database products.
The Quick Install Program supports three types of HACMP cluster configurations
with Oracle: Mutual Takeover, Rotating Standby and Hot Standby. It assumes the
user has first set up the cluster hardware correctly, including connecting shared
disks, connecting network adapters and RS232 cables, according to instructions
in the HACMP for AIX Installation Guide. It also assumes the user has installed
AIX 4.2 or higher and HACMP 4.2 or higher. From there, the Quick Install
Program configures network adapters, creates shared volume groups and logical
volumes, configures the HACMP cluster, and installs the Oracle products.
The first three chapters of the redbook provide general information about
relational databases and Oracle, and the chapters beyond that directly deal with
planning for, preparing for, and running the Quick Install Program. Some
knowledge of AIX, the RISC System/6000, and HACMP for AIX is assumed.
The Team That Wrote This Redbook
This redbook was produced by a team of consultants working at the International
Technical Support Organization, Austin Center.
Scott Brudner
John Neidhart
Neidhart Consulting Services of Austin, Texas.
Neidhart Consulting is a full-service provider for computer information systems.
Consulting services for programming, networking, systems/process analysis and
design, as well as hardware installation, and repairs are available. (512)
837-1491
Thanks to the following people for their invaluable contributions to this project:
Dave Thiessen
International Technical Support Organization, Austin Center
Marcus Brewer, Editor
International Technical Support Organization, Austin Center
Roger Feigelson, Product Line Marketing Manager
IBM Products Division, Oracle Corporation
Redwood Shores, CA
Comments Welcome
Your comments are important to us!
© Copyright IBM Corp. 1997
xiii
xii
ORACLE Cluster POWERsolution Guide
Tables
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
© Copyright IBM Corp. 1997
Hardware Requirements for Oracle Product Installation on a Single Server . .
Software Requirements for Oracle Product Installation on a Single Server. . .
Prerequisites for PL/SQL Release 2.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Prerequisites for SQL*Plus Release 3.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Prerequisites for SQL*Net Release 2.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Prerequisites for Oracle TCP/IP Adapter Release 2.3 . . . . . . . . . . . . . . . . . . .
Disk Sizing Sheet for Oracle Server Products . . . . . . . . . . . . . . . . . . . . . . . . .
Disk Sizing Sheet for Oracle Networking Products . . . . . . . . . . . . . . . . . . . . .
Disk Sizing Sheet for Oracle Tools Products . . . . . . . . . . . . . . . . . . . . . . . . . .
Disk Sizing Sheet for Oracle CASE Products . . . . . . . . . . . . . . . . . . . . . . . . .
First-Phase Disk Sizing Sheet for Oracle Precompiler Products . . . . . . . . . . .
Second-Phase Disk Sizing Sheet for Oracle Precompiler Products . . . . . . . .
Total Disk Sizing Summary Sheet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Disk Sizing Result Sheet for Oracle Server Products . . . . . . . . . . . . . . . . . . .
Disk Sizing Result Sheet or Oracle Tools Products . . . . . . . . . . . . . . . . . . . . .
Disk Sizing Result Sheet for Oracle Networking Products . . . . . . . . . . . . . . .
Total Disk Sizing Result Sheet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Common Errors During the Installation Stage . . . . . . . . . . . . . . . . . . . . . . . . .
Node1 Adapters for Mutual Takeover Configuration . . . . . . . . . . . . . . . . . . . .
Node2 Adapters for Mutual Takeover Configuration . . . . . . . . . . . . . . . . . . . .
Database Logical Volumes (Node 1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Database Logical Volumes (Node 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Node 1 Adapters for Rotating Configuration . . . . . . . . . . . . . . . . . . . . . . . . . .
Node 2 Adapters for Rotating Configuration . . . . . . . . . . . . . . . . . . . . . . . . . .
Database Logical Volumes for Both Nodes in Rotating Configuration . . . . . . .
Node1 Adapters for Hot Standby Configuration . . . . . . . . . . . . . . . . . . . . . . . .
Node2 Adapters for Hot Standby Configuration . . . . . . . . . . . . . . . . . . . . . . . .
Database Logical Volumes for Both Nodes in Hot Standby Configuration. . . .
17
17
18
18
18
18
20
20
22
24
25
26
27
34
35
36
37
85
89
89
90
90
92
92
93
94
94
95
xi
x
ORACLE Cluster POWERsolution Guide
Figures
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
© Copyright IBM Corp. 1997
Relationship between Two Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Logical Structure of a Typical Oracle Database . . . . . . . . . . . . . . . . . . . . . . . . . 8
Oracle Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Cluster Configuration for Rotating Resources . . . . . . . . . . . . . . . . . . . . . . . . . 30
Cluster Configuration for Hot Standby Resources . . . . . . . . . . . . . . . . . . . . . . 31
Cluster Configuration for mutual takeover Resources . . . . . . . . . . . . . . . . . . . 32
.rhosts File from node1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
/etc/hosts file from node1 Preinstallation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Adapter Test on node1_standby . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Logical Volumes for Oracle 7 Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Oracle Processes Running on Node with Database . . . . . . . . . . . . . . . . . . . . 84
Entry Needed in /etc/services for the Listener . . . . . . . . . . . . . . . . . . . . . . . . . 97
/home/oracle/.profile File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
/etc/oratab File for Rotating or Hot Standby Resources. . . . . . . . . . . . . . . . . . 98
listener.ora File for Rotating or Hot Standby Resources . . . . . . . . . . . . . . . . . 99
tnsnames.ora File for Rotating or Hot Standby Resources . . . . . . . . . . . . . . . 99
Entry Needed in /etc/services File for the Listener. . . . . . . . . . . . . . . . . . . . . 101
/home/oracle/.profile File on First Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
/home/oracle/.profile File on Second Node . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Node 1 oratab Entry for the Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
Node 2 oratab Entry for the Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
oratab Entries for Both Databases Running on One Node . . . . . . . . . . . . . . 103
listener.ora File for Mutual Takeover Resources, Node 1 . . . . . . . . . . . . . . . 103
listener.ora File for Mutual Takeover Resources, Node 2 . . . . . . . . . . . . . . . 104
tnsnames.ora File for Mutual Takeover Resources . . . . . . . . . . . . . . . . . . . . 104
ix
C.1.5 File: //node1_scripts/tnsnames.ora . . . . . . . . . . . . . . . . . . . . . . . . . . . .104
Appendix D. Special Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .105
Appendix E. Related Publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .107
E.1 International Technical Support Organization Publications . . . . . . . . . . . . . .107
E.2 Redbooks on CD-ROMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .107
E.3 Other Publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .107
How To Get ITSO Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
How IBM Employees Can Get ITSO Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . .109
How Customers Can Get ITSO Redbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .110
IBM Redbook Order Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .111
List of Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
ITSO Redbook Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
viii
ORACLE Cluster POWERsolution Guide
6.29 Specify National Language Support (NLS) . . . . . . . . . . . . . . . . . . .
6.30 Relink Executables? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.31 Location of Postinstallation File . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.32 Online Help and Documentation . . . . . . . . . . . . . . . . . . . . . . . . . .
6.33 Products Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.34 Official Hostname . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.35 TCP Service Port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.36 Passwords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.37 Group Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.38 More Than One Instance? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.39 Choose Storage Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.40 Select the Character Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.41 Enter the Password for the SYSTEM User and the SYS User . . . .
6.42 Do You Want To Set the Passwords for DBA and Operator? . . . . .
6.43 MTS Configured and the SQL*Net Listener Automatically Started?
6.44 Database Domain Name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.45 Name Control Files and Tablespaces . . . . . . . . . . . . . . . . . . . . . .
6.45.1 Name and Size Tablespaces and Logs: . . . . . . . . . . . . . . . . .
6.46 Optional Product Documentation . . . . . . . . . . . . . . . . . . . . . . . . . .
6.47 Post-Installation Program. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.48 Copying and Linking Necessary Files . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
61
62
62
63
63
64
65
65
65
65
66
66
66
66
66
67
67
69
78
78
78
Chapter 7. Post-Installation of the Oracle Server Database
7.1 Verify Database Security and Integrity . . . . . . . . . . . . . . . .
7.2 Create Oracle Server User Logins . . . . . . . . . . . . . . . . . . .
7.3 Check your Filesystems. . . . . . . . . . . . . . . . . . . . . . . . . . .
7.4 Testing Cluster Synchronization . . . . . . . . . . . . . . . . . . . .
7.5 HACMP Cluster Services . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
81
81
81
81
83
83
Chapter 8. Common Errors, Problems and Recommendations
8.1 Common Errors and Problems. . . . . . . . . . . . . . . . . . . . . . . . .
8.1.1 Errors and Problems During the Installation Process . . . .
8.1.2 SQL*Net Errors and Problems . . . . . . . . . . . . . . . . . . . . .
8.2 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
85
85
85
86
87
Appendix A. Quick Installation Worksheets . . . . . . . . . . . . . . . . . . . . . . . . . .89
A.1 Mutual Takeover Worksheet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89
A.2 Rotating Standby Worksheet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92
A.3 Hot Standby Worksheet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .94
Appendix B. Configuration Files for Rotating & Hot Standby Resources .97
B.1 Customized Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97
B.1.1 File: /etc/services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97
B.1.2 File: /home/oracle/.profile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97
B.1.3 File: /etc/oratab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
B.1.4 File: //node1_scripts/listener.ora . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
B.1.5 File: /node1_scripts/tnsnames.ora . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99
Appendix C. Oracle Configuration Files for Mutual Takeover Resources 101
C.1 Customized Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101
C.1.1 File: /etc/services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101
C.1.2 File: /home/oracle/.profile. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101
C.1.3 File: /etc/oratab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102
C.1.4 File: /node1_scripts/listener.ora . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103
vii
ORACLE Cluster POWERsolution Guide
3.2.2
3.3 Disk
3.3.1
3.3.2
3.3.3
Requirements for Oracle Optional Products .
Space and Memory Sizing for Oracle . . . . . .
Required Calculations . . . . . . . . . . . . . . . . .
Disk Space Requirements . . . . . . . . . . . . . .
Final Calculation of Disk Requirement . . . . .
Chapter 4. HACMP Cluster Configurations
4.1 Rotating Standby Configuration . . . . . . . .
4.2 Hot Standby Configuration . . . . . . . . . . . .
4.2.1 Hot Standby Cluster Description . . .
4.3 Mutual Takeover Configuration . . . . . . . .
vi
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
17
19
19
19
27
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
29
30
31
31
32
Chapter 5. Quick Install Program Preparation . . .
5.1 Choosing Oracle Products to Install . . . . . . . . . .
5.2 Disk Space Sizing for Selected Oracle Products
5.3 Copying Files From Floppy Disk . . . . . . . . . . . .
5.4 Installation Worksheet . . . . . . . . . . . . . . . . . . . .
5.5 Assign Network Adapters. . . . . . . . . . . . . . . . . .
5.6 Update /.rhosts and /etc/hosts . . . . . . . . . . . . . .
5.6.1 Update .rhosts on Both Nodes . . . . . . . . . .
5.6.2 Update /etc/hosts on Both Nodes. . . . . . . .
5.7 Find Available Group & User ID Numbers . . . . .
5.8 Find the Correct Terminal Emulation . . . . . . . . .
5.9 Creation of Logical Volumes . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
33
33
33
37
37
37
37
38
38
39
39
40
Chapter 6. Running the Quick Install Program . . . . . . . . . . . .
6.1 Declaring Node Number . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.2 Declaring Failover Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3 Choose TTY Port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.4 Select Network Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.5 Adapter Configuration Screen . . . . . . . . . . . . . . . . . . . . . . . .
6.6 Information Menus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.7 Change Default Names . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.8 Create DBA Group. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.9 Create Oracle User Account . . . . . . . . . . . . . . . . . . . . . . . . .
6.10 Creating Filesystems, Directories, and Setting Permissions.
6.11 Setting Oracle User Environment. . . . . . . . . . . . . . . . . . . . .
6.12 Creating Volume Groups . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.13 Creation of Logical Volumes . . . . . . . . . . . . . . . . . . . . . . . .
6.14 Mount CD-ROM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.15 Linking Oracle 7 Files to Staging Area . . . . . . . . . . . . . . . . .
6.16 Running The Preinstallation Script . . . . . . . . . . . . . . . . . . . .
6.17 Oracle 7 Welcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.18 Enter Installer Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.19 Installation Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.20 Entering the Mount Point . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.21 Create Database Objects. . . . . . . . . . . . . . . . . . . . . . . . . . .
6.22 Information Prompts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.23 OPS Install . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.24 Preinstallation Preparation. . . . . . . . . . . . . . . . . . . . . . . . . .
6.25 Install Products on All Nodes? . . . . . . . . . . . . . . . . . . . . . . .
6.26 Installation Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.27 Source Staging Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.28 ORACLE_SID Variable . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
41
41
41
42
42
42
47
48
48
49
50
51
51
53
54
55
55
56
56
57
57
58
58
58
59
59
60
61
61
ORACLE Cluster POWERsolution Guide
.
.
.
.
.
.
.
.
.
.
Contents
Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .v
Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
The Team That Wrote This Redbook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Comments Welcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Chapter 1. Relational Database Concepts .
1.1 Relational Databases . . . . . . . . . . . . . . .
1.2 Database Structure . . . . . . . . . . . . . . . . .
1.3 Structured Query Language (SQL) . . . . .
1.3.1 SQL Statements . . . . . . . . . . . . . . .
1.3.2 Context Areas and Cursors . . . . . . .
1.3.3 Statement Parsing. . . . . . . . . . . . . .
1.4 Client/Server Architecture . . . . . . . . . . . .
1.4.1 Client/Server Connection. . . . . . . . .
1.4.2 Network Connections . . . . . . . . . . .
1.4.3 Shared-Memory Connections . . . . .
1.5 Filesystems and Raw Logical Volumes . .
1.5.1 Filesystems . . . . . . . . . . . . . . . . . . .
1.5.2 Raw Logical Volumes . . . . . . . . . . .
© Copyright IBM Corp. 1997
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
1
2
2
2
3
3
4
4
4
4
5
5
5
Chapter 2. Oracle Concepts . . . . . . . . . . . . . . . . . . . . . . .
2.1 Oracle Files. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.1 Data Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.2 Redo Log Files . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.3 Control File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.1.4 The init.ora File . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2 Oracle Memory Structure . . . . . . . . . . . . . . . . . . . . . . .
2.2.1 System Global Area or Shared Global Area (SGA)
2.2.2 Program Global Area (PGA) . . . . . . . . . . . . . . . . .
2.3 Oracle Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3.1 Database Writer (DBWR) . . . . . . . . . . . . . . . . . . .
2.3.2 Log Writer (LGWR) . . . . . . . . . . . . . . . . . . . . . . . .
2.3.3 System Monitor (SMON) . . . . . . . . . . . . . . . . . . . .
2.3.4 Process Monitor (PMON). . . . . . . . . . . . . . . . . . . .
2.3.5 Checkpoint (CKPT) . . . . . . . . . . . . . . . . . . . . . . . .
2.3.6 Recover (RECO) Process . . . . . . . . . . . . . . . . . . .
2.3.7 Archiver (ARCH) Process . . . . . . . . . . . . . . . . . . .
2.3.8 Parallel Server Lock Processes (LCKn) . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7
7
7
9
9
10
10
10
10
11
13
13
13
13
13
14
14
14
Chapter 3. Database Planning, System Requirements, and Sizing . . .
3.1 Planning Your Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.1 First Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.2 Database Application Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.3 Database Size. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1.4 File Locations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2 System Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2.1 Requirements for Oracle Product Installation on a Single Server.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
15
15
15
15
16
17
17
17
v
Take Note!
Before using this information and the product it supports, be sure to read the general
information in Appendix D, “Special Notices” on page 105
First Edition (December 1997)
This edition applies to IBM Interactive Network Dispatcher Version 1.2 for AIX.
Comments may be addressed to:
IBM Corporation, International Technical Support Organization
Dept. JN9B Building 045 Internal Zip 2834
11400 Burnet Road
Austin, Texas 78758-3493
When you send information to IBM, you grant IBM a non-exclusive right to use or distribute
the information in any way it believes appropriate without incurring any obligation to you.
© Copyright International Business Machines Corporation 1997. All rights reserved
Note to U.S Government Users – Documentation related to restricted rights – Use, duplication or disclosure is
subject to restrictions set forth in GSA ADP Schedule Contract with IBM Corp.
SG24-2019-00
International Technical Support Organization
Oracle Cluster POWERsolution Guide
December 1997
SG24-2019-00
Oracle Cluster POWERsolution Guide
December 1997
International Technical Support Organization
Austin Center
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement