Performing Backups and Recovering Your Database

Log in to Enterprise Manager Database Console by opening your browser and entering the following URL:

http://raclinux1.us.oracle.com:1158/em

Enter sys/oracle as SYSDBA and click Login.

The Oracle Database Home page allows you to view the current state of the database by displaying a series of metrics that portray the overall health of the database. The Oracle Database Home page provides a launch point for the database status and administration and configuration of the database environment. It contains four pages via subtabs with each page displaying subsections. Click on the Maintenance tab.

Under Backup/Recovery Settings, click Recovery Settings.

Scroll down to the Flash Recovery Area section.

Enter +MY_DG1 for the Flash Recovery Area Location and 2 GB for the Flash Recovery Area Size. Then scroll up to the top of the page.

Click Apply.

The Flash Recovery Area has been set. Now you need to put the database in archivelog mode. Scroll down to the Media Recovery section.

Click the ARCHIVELOG Mode checkbox.

Scroll down to the Flash Recovery Area region of the page and click Enable Flashback Database. Change the Flashback Retention Time to 6 hours, then click Apply.

Click Yes to restart the database.

Enter your Cluster Credentials as oracle/oracle and Database Credentials as sys/oracle as SYSDBA. Check the box to Save as Preferred Credentials and then click Continue.

Your database is being restarted. Wait about 10 minutes and click Refresh.

Once you click refresh, you will receive the Enterprise Manager Login Page. Enter sys/oracle as SYSDBA and click Login.

Select the Maintenance tab.

Select Backup Settings in the Backup/Recovery Settings region.

Click on the Policy tab.

Check Enable block change tracking for faster incremental backups. For the name of the Block Change Tracking File enter +MY_DG1. Scroll to the bottom of the page.

Verify the host credentials are set. Check the Save as Preferred Credential option, then click OK.

On the Maintenance page, under Backup/Recovery Settings, click Backup Settings.

Change the Disk Backup Type to Image Copy. Then, to make sure the disk backup interface is working, click Test Disk Backup.

You will see the progress window.

After one minute or less, you see that your Disk Backup Test was successful. Scroll down.

Make sure your Host Credentials are set to oracle/oracle and click OK to save the settings.

Click Schedule Backup in the Backup/Recovery section on the Maintenance page.

To minimize the administrative tasks of performing backups, you will use the Oracle Suggested Backup method. This method schedules a daily backup job that backs up the entire database using the backup settings configured in the previous section.

Click Schedule Oracle-Suggested Backup for the Backup Strategy.

Make sure Disk is selected, then click Next.

Click Next.

Set the time to 5 minutes after the current time, then click Next. Note: you can identify the current time if you look at the upper right corner of VMware.

Click Submit Job to schedule the backup.

Your backup job has been submitted. Click View Job .

Click on Reload or Refresh in your browser until the Page Refreshed time shows a time after the scheduled start time of the backup job.

Once your job starts, you can see its progress. Click on the Backup log to see more details about this execution.

Scroll down to see the details.

After the backup has completed successfully in about 7 minutes, click the Database tab at the top of the page.

On the Maintenance page, in the Backup/Recovery region, click Backup Reports.

A list of backup jobs currently scheduled within the database is displayed. Click on the entry for the full backup.

The report shows summary information for the backup job, such as the number of files backed up, the size of the backup, and the files contained within it. Scroll down to view the Inputs and Outputs sections.

These sections of the report provide more detailed information about the files that were backed up by this job. You can see the file numbers, file size, backup type, number of corrupt blocks and checkpoint SCN information. Scroll down to view the next three sections of the reports.

Now you see information regarding the Backup Sets, Backup Pieces, and Image Copies. Scroll back up to the top of the page.

Click the Cluster Database breadcrumb.

From the Maintenance tab, under Backup/Recovery Settings, click Recovery Settings.

Scroll down to the Flash Recovery section. You will see a pie chart which contains the usage of the flash recovery area. Notice that because you just performed a backup, the Image Copy is taking up the most space.

Scroll up to the top of the page and click the Database Cluster breadcrumb.

From the Maintenance tab, scroll down to the bottom of the page and click Jobs under Related Links.

Select RMAN Script from the Create Job pulldown list and click Go.

For the job name, enter DAILY_ARCH_BACKUP. Enter a description for this job, such as 'Backup archived log files in the flash recovery area'. For the RMAN script, you can either choose to have a saved script executed by typing in @/home/oracle/wkdir/backup_arch.rman or you can type in the RMAN script to run, such as the one listed here:

run {
  backup archivelog all delete input format '+MY_DG2';
}

When finished typing in the information, click Add to specify a database target.

Click the Search button.

Select RACDB and click Add.

On the Create 'RMAN Script' Job page, scroll down to the Host and Database Credentials region. Select Override Preferred Credentials and enter oracle/oracle for the Host User Name and Password and confirm the password. Enter sys/oracle for the Database User Name and password and confirm the password. Change the Database Role to SYSDBA then click the Schedule tab.

Select Later for the Start period and set the time to 11:30 pm. In the Repeat region, select Interval and set the Frequency to 1 Days. Then click Submit.

The job is then created and scheduled. To check the current status of the scheduled job, click the Database tab.

On the Cluster Database: RACDB page, scroll down to the Job Activity region and click the linked number under the heading 'Submitted to the Cluster Database'.

On the Job Activity page, you can see that the DAILY_ARCH_BACKUP job is scheduled to run at a later time. Click the Database link at the bottom of the page..

Open a terminal window. Invoke SQL*Plus and connect as the HR user. Query the EMPLOYEES table as follows:

sqlplus /nolog
connect hr/hr
SELECT count(*) FROM hr.employees;
exit

In order to delete a data file, you need to take the tablespace offline, then delete it. Using Enterprise Manager, logged in as a SYSDBA user, click Administration tab.

To put the database in recovery mode, click the Perform Whole Database Recovery button.

Click Yes to shutdown the database.

The Recovery Wizard page is displayed. Wait about 8 minutes, then click Refresh.

Click Startup.

Enter oracle/oracle for the host credentials, then click Continue.

Select the RACBD1 instance and click Startup.

To specify the instance should be mounted, not opened, click Advanced Options.

Select Mount the database and click OK.

Click Yes to confirm.

Open a terminal window and use the following commands to verify the instance has been started and is MOUNTED, not OPEN:

sqlplus /nolog
connect sys/oracle as sysdba

SELECT * FROM V$INSTANCE;

Flashback the database to the restore point created before the HR.JOBS_HIST table was truncated. Enter the following command in SQL*Plus:

FLASHBACK DATABASE TO RESTORE POINT before_trunc;

When the FLASHBACK command completes, open the database using the following command:

ALTER DATABASE OPEN RESETLOGS;

After a few minutes, the database is opened. Enter the following commands to check the contents and structure of the HR.JOBS_HIST table. Notice that the rows are restored, but the recently added column no longer exists.

connect hr/hr

SELECT count(*) FROM hr.jobs_hist;
DESCRIBE hr.jobs_hist;
exit;

Log in to Enterprise Manager Database Console by opening your browser and entering the following URL:

http://raclinux1.us.oracle.com:1158/em

Enter sys/oracle as SYSDBA and click Login.

On the Cluster Database Home page, the current state of the database shows one instance is not available. To restart the second instance, scroll down to the bottom of the page.

Click the RACDB_RACDB2 link.

Click Startup.

Make sure the host and database credentials are filled in, then click OK.

Click Yes to start and open the instance.

Wait about 2 minutes for the instance to start.

After the instance starts, you see the following message displayed. Click the Cluster Database breadcrumb.

Now on the Home page, under the General heading, you see that both instances are available.