dump can back up an entire system or perform incremental backups, saving changed files only since the last backup. Options supported include specifying the size and length of storage media. The backup storage device is usually a tape device or another hard disk partition. SCSI tape devices labeled with prefix st with a number attached like st0. For the device on another system on a network – a hostname entered before the device name delimited with a colon like ananova.com:/dev/st0.
Commands
- diskdumpfmt: used to format tapes for use by dump
- diskdumpctl: registers dump partition with the system
- savecore: saves a vmcore file from the data in a dump partition
- crash tool: To read dumped cores
- add [arg]: Adds files or directories to the list of files to be extracted — such files tagged with * before names when listed with ls.
- extract: Extracts files and directories on the extraction list
dump utility uses dump levels to determine the degree to a backup system.
- –0 through -9: The dump level 0 is a full backup, i.e., copy file systems in their entirety. The dump levels above 0 perform incremental backups, backing up files and directories created or modified since the last lower-level backup. The dump level 1 backup only files changed since the last level 0 backup and dump level 2, in turn, will back up only files changed since the last level 1 backup, and so up to dump level 9. The default level is 9.
- -B records: specify the number of blocks in a volume overriding the end-of-media detection or length and density calculations that dump usually uses for multivolume dumps.
- -a: It lets dump bypass any tape length calculations and writes until an end-of-media indication is detected. It is recommended for most modern tape drives and is the default.
- -b blocksize: Specify the number of KB per dump record, create larger blocks, thus speeding up backups
- -h level: Files tagged with the user's nodump flag will not back up at or above the specified level. The default is 1, which will not backup the tagged file in incremental backups.
- -f file/device: Backs up the file system to a specified file or filenames separated by commas or device or remote device or file referenced with a preceding hostname.
- -k: uses Kerberos authentication to talk to remote tape servers
- -M file/device: Implements multivolume backup where files need to be greater than the Linux ext3 2 GB file size limit. The ‘file' written as a prefix to a numbered sequence from 001 like file001, file002, and so on.
- -n: notifies operators if backup needs operator attention
- -s feet: specifies tape length in feet. dump prompts for new tape when the length reached.
- -T date: Allows date specification instead of using /etc/dumpdates file
- -u: Records the backup or writes entry for a successful update in the /etc/dumpdates file
- -W: detects and displays the file system that needs to be backed up. The information is taken from the /etc/dumpdates and /etc/fstab files.
- -w: detects and display the file system that needs to be backed up, drawing only the information in /etc/fstab. It uses the /etc/fstab to report the file systems ready for backup
Files
- /etc/dumpdates: The file records and lists backups. It specifies the backups, file system, date, and dump level. This information is required to restore files from a specified backup.
- /etc/fstab: records the dump level, and backup frequency for each file system
Recovering Backups using restore command
It restores an entire file system or particular files or directories from a backup archive and copies them to the current working directory. It would generate sub-directories as needed.
mkfs /dev/hda2 mount /dve/hda2 /ananova cd /ananova restore -ivf /dev/tape
Options
- i: The interactive mode for restoring particular files and directories in a backup
- -C: checks backup by comparing files on a file system with those in a backup
- -R: Instructs restore to request a tap that is part of a multi-volume backup
- -r: Restores a file system
- -t: Lists the content of a backup or specified files in it
- -x: Extracts specified files or directories with its sub-directories from a backup. If no file or directory specified, the entire file system restored.
- -b blocksize: specifies block size other it dynamically determine it from the block device
- -f file/device: Runs on specified file or device; hostname specified on remote devices
- -F script: Runs a script at the beginning of the restore
- -k: uses Kerberos authentication to remote devices
- -h: Extracts only the specified directories without their sub-directories
- -M files/devices: Restores from multivolume backups where ‘file' prefixes on a numbered sequence, file001, file002.
- -N: Displays the names of files and directories without extracting them
- -T directory: Specifies the directory to use for the storage of temporary files, the default value is /tmp
- -v: The verbose mode
- -y: Suppresses the query, allowing restore to continue automatically.