Backup and Recovery - Oracle Exadata Recipes: A Problem-Solution Approach - page 192

Solution
In this recipe, you will learn how to do the following:

• Display available space in your current compute node volume groups

• Create an additional logical volume to store your backups

• Create a file system on your backup logical volume and mount the file system

• Mount the file system

/ and / u01 file systems • Create LVM snapshots on your

• Mount your snapshot volumes

• Back up the contents of the snapshot file systems to your backup volume

•
Begin by logging in to the compute node server that you wish to back up as root and perform the following steps:

Unmount and remove your snapshot file systems

Run vgdisplay to display your free storage capacity: 1.

[root@cm01dbm01 ~]# vgdisplay | egrep '(VGExaDb|Alloc PE|Free PE)'
VG Name VGExaDb
Alloc PE / Size 39424 / 154.00 GB
Free PE / Size 103327 / 403.62 GB
[root@cm01dbm01 ~]#

Unless you have previously extended VGExaDb , you should see about 400 GB of free space
for the volume group.

■ the /u01 file system on the exadata Compute nodes is created on a 100 GB logical volume. the underlying
volume group for this logical volume is used for both the /u01 and root file system, as well as a swap partition, and
is 600 GB in size. this means that you will typically have a large amount of unallocated disk storage to use for either
extending the default volumes or creating new logical volumes.

Note

Since we want to back up our root and /u01 file systems, which currently use 154 GB of
space from the VGExaDb volume group, we will create a new logical volume in this volume
group. In this example, we're calling this backup :

2.

[root@cm01dbm01 ~]# lvcreate -L 154G -n /dev/VGExaDb/backup
Logical volume "backup" created
[root@cm01dbm01 ~]#

When your backup volume group is created, format it using mkfs.ext3 using a 4 KB block
size with the following command:

3.

[root@cm01dbm01 ~]# mkfs.ext3 -m 0 -b 4096 /dev/VGExaDb/backup
mke2fs 1.39 (29-May-2006)
Filesystem label=