Using PXE to batch push Oracle 11gR2 RAC environment from bare metal

The PXE method is used to batch push the Complete Oracle 11gR2 RAC environment from bare metal, and the Shell script is used to deploy 32 Oracle11gR2 RAC backup and recovery cases on the launched RAC nodes in batches.

Consultant for ARP Project Implementation of the Chinese Emy of sciences, Tang Bo, a member of the Shanghai Oracle user group

Summary

In the first half of this article, we will introduce how to use the PXE push terminal host to batch Push three PXE In the bare metal status to the Oracle 11gR2 RAC environment. In each Oracle 11gR2 RAC environment, three hosts are automatically installed and configured: one shared disk host and two node hosts. Each of the two node hosts in the Oracle 11gR2 RAC environment has been automatically attached to the shared disk hosts in the environment. Each of the two node hosts is equipped with a distributed virtual tape library and all operating system configurations, including vip, scan-vip, ntp, various rpm packages, and operating system parameters.
The second half of this article introduces how to reuse the above PXE push terminal host as the deployment platform, and deploy 32 Oracle11gR2 RAC backup and recovery case scripts in each set of Oracle 11gR2 RAC environments. The script is executed and deployed in batches based on the network. The script can be downloaded to each Oracle11gR2 RAC node host for local operation. The script can simulate and evaluate the recovery results of each Backup recovery scenario in each environment.

All original programs mentioned in this article provide download and md5sum files.

:

------------------------------------------ Split line ------------------------------------------

FTP address: ftp://ftp1.bkjia.com

Username: ftp1.bkjia.com

Password: www.bkjia.com

In, LinuxIDC.com \ February \ The PXE method was used to batch push the Oracle 11gR2 RAC environment from bare metal

For the download method, see

Or to Baidu Network Disk download: http://pan.baidu.com/share/link? Consumer id = 3474287943 & uk = 487907638

------------------------------------------ Split line ------------------------------------------

Readers should not use these procedures for commercial purposes, and the author shall not assume any legal responsibility for any consequences arising from the use of these procedures.

 

Directory

1. PXE Overview
1.1 PXE Overview
1.2 IP address planning
2. Build a PXE pushing host
2.1 build a PXE pushing host
2.2 download the installation software to be used for pushing
2.3 collect MAC addresses from the pushing host of PXE
3. PXE push HOST: Oracle 11gR2 RAC environment shared disk host
4. PXE push HOST: Oracle 11gR2 RAC environment node host
4.1 shared storage
4.2 distributed virtual tape library
4.3 Time Synchronization
4.4 grid users and oracle users
4.5 ssh equivalence script
4.6 others
4.7 Installation
5. Use Shell scripts to deploy 32 Oracle11gR2 RAC backup and recovery cases in batches on the released nodes
5.1 Working Principle of script simulation in Backup recovery scenarios
5.2 install Backup recovery case simulation script on the PXE push host
5.3 deploy Backup recovery case simulation script on the node host in Oracle 11gR2 RAC Environment
5.3.1 prepare an IP address list on the PXE push host
5.3.2 initialize the experiment script on the PXE pushing host
6. Oracle11gR2 RAC environment "dd" and "rever", initialize the environment through "rever" before a specific experiment
7. Oracle11gR2 RAC Backup recovery case 1: bulk simulation and recovery points for full recovery scenarios
7.1 1a_users tablespace damaged online
7.2 1b _ The tablespace of the deprecated user is damaged.
7.3 1c _ read-only user tablespace corruption
7.4 1d_users tablespace hot standby
7.5 3_system tablespace offline corruption
7.6 4_tbsocp05_test the tablespace not backed up is corrupted.
8. Oracle11gR2 RAC Backup recovery case 2: Batch simulation and recovery points in Incomplete recovery scenarios
8.1 5 _ time-based Incomplete recovery
8.2 6a _ Incomplete recovery based on log serial numbers
8.3 6b _ incomplete restore Based on cancel
8.4 7a _ current control file and data file intact _ all log files damaged _ normal shutdown Incomplete recovery _ No data loss _ No backup required
8.5 7b _ current control file and data file intact _ all log files damaged _ abnormal shutdown Incomplete recovery
8.6 8a _ damage to current control file _ Incomplete recovery _ Use of binary backup of control file _ No data loss _ No backup required
8.7 8b _ damage to current control file _ full recovery _ Use Control File Script _ No backup required
8.8 9a _ damage to current control file _ Incomplete recovery of table space of offline user _ Use of binary backup of control file _ No data loss _ No backup required
8.9 9b_damage to the current control file _ completely recovering the table space of the offline user _ using the control File Script _ No backup required
8.10 9c _ damage to current control file _ Incomplete recovery of read-only user tablespace _ Use of binary backup of control file _ No data loss _ No backup required
8.11 9d _ damage to current control file _ full recovery of read-only user tablespace _ Use of Control File Script _ No backup required
8.12 10a _ damage to current control file _ Incomplete recovery of user tablespace during Backup _ Use of binary backup of control file _ No data loss _ No backup required
8.13 10b _ damage to current control file _ full recovery of user tablespace during Backup _ Use of Control File Script _ No backup required
8.14 10c _ damage to current control file _ Incomplete recovery of read-only user tablespace during Backup _ Use of binary backup of control file _ No data loss _ No backup required
8.15 10d _ damage to current control file _ full recovery of read-only user tablespace during Backup _ Use of Control File Script _ No backup required
9. Oracle11gR2 RAC Backup recovery case 3: Advanced Incomplete recovery scenario batch simulation and recovery key points
9.1 11a _ the current control file and log file are all damaged _ Incomplete recovery _ binary backup of the Control File
9.2 11b _ the current control file and log file are all damaged _ Incomplete recovery _ use the control File Script
9.3 12a _ current control file and log file are all damaged _ backup current line user tablespace is not completely restored _ Use Control File binary backup
9.4 12b _ current control file and log file are all damaged _ backup current line user tablespace is not completely restored _ Use Control File Script
9.5 12c _ the current control file and log file are all damaged _ the read-only user tablespace is not completely restored during Backup _ use the control file binary backup
9.6 12d _ the current control file and log file are all damaged _ the read-only user tablespace is not completely restored during Backup _ use the control File Script
9.7 13a _ damage to current control file _ creating tbsocp05_test2 tablespace Incomplete recovery _ using control file binary backup _ No data loss _ No backup required
9.8 13b _ damage to current control file _ create tbsocp05_test3 tablespace full recovery _ Use Control File Script _ No backup required
9.9 The tablespace deleted cannot be completely restored.
9.10 15_incarnation not completely restored
Summary

Body

1. PXE Overview

1.1 PXE Overview

PXE (Pre-boot Execution Environment) is an Intel-designed protocol that enables bare metal to be started over the network and installed on the operating system. The protocol can be divided into push end and push end.
The PXE push host is a standard Linux server that runs dhcpd, tftpd, nfs, dns, vsftpd, and httpd collaboratively.
PXE is activated by the push program from the ROM of the NIC. During computer boot, the BIOS transfers PXE to the memory for execution by the push end program and displays the command menu. After user selection, the PXE push end downloads the operating system of the host on the PXE push end to run locally through the network, and then starts the kernel. After the PXE is started, the push end actively downloads the installer and Installation option files from the PXE push end host to complete subsequent installation steps.
1.2 IP address planning

In order to clearly describe the following, we set the network segment to 192.168.0.0/24. The operating system uses 6 series RHEL and 5 Series OEL respectively. This article also applies if the steps for using other 6-series Enterprise Linux and 5-series Enterprise Linux (such as CentOS, RHEL, or OEL, including their 32-bit or 64-bit distributions) are similar.

Host	Operating System	Internet IP Address	Intranet IP Address	Host Name
PXE push host	RedHat Enterprise Linux 6.1 x86_64 (or another 6-series Enterprise Linux and 5-series Enterprise Linux)	192.168.0.254	None	Server1.example.com
PXE: Oracle 11gR2 RAC environment shared disk host	RedHat Enterprise Linux 6.1 (or another 6-series Enterprise Linux)	192.168.0.50 + x (The value range of x is 1 to 39)	172.31.118.50 + x (The value range of x is 1 to 39)	Station50 + x.example.com (The value range of x is 1 to 39)
Oracle 11gR2 RAC environment node host (first server) of PXE push host)	Oracle Enterprise Linux 5.4 (or other 5-series Enterprise Linux)	IP: 192.168.0.x VIP: 192.168.0.200 + x SCAN-VIP1: 192.168.0.100 + x SCAN-VIP2: 192.168.0.150 + x (The value range of x is 1 to 39)	172.31.118.x (The value range of x is 1 to 39)	Stationx.example.com (The value range of x is 1 to 39)
Oracle 11gR2 RAC environment node host (Second) of PXE push host)	Same as above	IP: 192.168.0.1 + x VIP: 192.168.0.201 + x SCAN-VIP3: 192.168.0.101 + x (The value range of x is 1 to 39)	172.31.118.1 + x (The value range of x is 1 to 39)	Station1 + x.example.com (The value range of x is 1 to 39)

During the push, the above IP addresses are planned through the MAC fixed configuration in the dhcpd configuration file on the PXE pushing host. The above VIP plan is implemented through the "% post" programming in the Kickstart configuration file on the PXE pushing host. The above three SCAN-VIP plans are implemented through the dns server configuration file of the PXE pushing host and the "% post" programming in the Kickstart configuration file on the PXE pushing host.

2. Build a PXE pushing host

2.1 build a PXE pushing host

As mentioned above, the PXE push host is a standard Linux server that runs dhcpd, tftpd, nfs, dns, vsftpd, and httpd collaboratively. Since this article is based on Oracle technology, it is not intended to detail the specific configurations of these servers. To help readers build the PXE pushing host, the author compiled a one-step setup program "bcp-install ".

"Bcp-install" has only one function: it helps you configure it within three minutes: a standard Linux server that runs dhcpd, tftpd, nfs, dns, vsftpd, and httpd collaboratively. This program is actually a shell-programmed self-extracting installation package. Therefore, it can be run on both 32-bit and 64-bit operating systems, 6-series Enterprise Linux, and 5-series Enterprise Linux.

Before running the setup program, you must install an Enterprise Linux host on your own. Only one Nic is required for this host. The IP address and host name should be set according to the 1.2 IP address plan. During the installation of this host, the "/usr" space is not less than 20 GB, and the "/var" space is not less than 15 GB (if no partition is installed, as long as the "/" is not less than 45 GB) and pick packages on the ftp server without special requirements. To make the script "bcp-install" run smoothly, you need to install all the content of the entire Disc used when the host operating system is installed, copy the complete directory structure directly to/var/ftp/pub. Then run "bcp-install ":

[Root @ server1 ~] #./Bcp-install ......

Shocould copy cdrom/DVDROM of linux installation into/var/ftp/pub :( y/n)

Make sure that all the content contained in the entire Disc used to install the operating system of the host is copied directly to "/var/ftp/pub/" and y is selected to continue.

So far, the establishment of the PXE pushing host is complete. If you need to know the specific configuration of "bcp-install", open the corresponding configuration files of dhcpd, tftpd, nfs, dns, vsftpd, and httpd on this machine, with the help of Linux-related knowledge, you can simply read it.

For more details, please continue to read the highlights on the next page:

Recommended reading:

Detailed description of the installation process of Oracle 11g Database in Linux

How to install Oracle 11g R2 single-instance database on CentOS 5.6

To install Oracle Clusterware on an Oracle vmvm

Install Oracle 11 GB single-instance database on Linux under vmvm

1 2 3 4 5 Next Page