ufsdump & ufsrestore

작성자

작성시간

2003-12-31 17:25:44

The ufsdump/ufsrestore command

ufsdump는 tar와 함께 가장 널리 쓰이는 Backup command이다. Backup이라는 측면에서만 본다면 tar보다 여러가지 면에서 유용한 command이다. tar와 비교를 해보면 일단, ufsdump는 multi-volume backup 이 가능하다. 따라서 Tape의 사용효율을 높일 수가 있다. 또, incremental backup을 기본적으로 지원하기 때문에, 정기적인 backup 프로그램으로는 대부분 ufsdump를 사용한다.
tar와 달리 ufsdump는 FileSystem단위로 Backup을 하기 때문에, 전체 backup을 하기 위해서는 각각의 FileSystem을 다 백업 해야 한다.

- backup
ufsdump의 사용법은 아래와 같다.
ufsdump [option] [argument] filesystem

option
0-9 Specify the dump level option. 0 is the lowest level(called the epoch level or a full dump), and level 9is the highest level.
u update the dump recrd(/etc/dumpdates) with the date and dump level of this file system backup
c Set the blocking factor to 126 for all catridge tape drives. This causes the dump to write 63-Kbyte records instead of the default 32-Kbyte records.
a Create an on-line archive of the file names dumped to tape.
f Specify the device to which the files are written. It requires an argument that is the device name.
v Verifies data on tape against data on file system.

argument
여기서는 target이 된다. 기본적으로 target은 /dev/rmt/0으로 되어 있지만, remote backup이나, incremental backup을 위해서는 지정이 필요하다.

예 ) file system 1개를 tape 1번째 volume에 backup하는 예.
e250 [ root{63} ~ ] df -k
파일시스템 K바이트 사용 가용 용량 설치지점
/dev/dsk/c0t8d0s0 96975 10 87268 1% /test_fs
e250 [ zeronine{5} ~ ] ls /test_fs
총 18 16 lost+found/ 0 nice 0 zero
0 good 0 min 2 test
e250 [ root{64} ~ ] ufsdump 0uf /dev/rmt/0 /dev/dsk/c0t8d0s0
DUMP: Writing 32 Kilobyte records
DUMP: Date of this level 0 dump: 2001년 4월 30일 월요일 오전 11시 45분 27초
DUMP: Date of last level 0 dump: the epoch
DUMP: Dumping /dev/rdsk/c0t8d0s0 (e250:/test_fs) to /dev/rmt/0.
DUMP: Mapping (Pass I) [regular files]
DUMP: Mapping (Pass II) [directories]
DUMP: Estimated 132 blocks (66KB).
DUMP: Dumping (Pass III) [directories]
DUMP: Dumping (Pass IV) [regular files]
DUMP: Tape rewinding
DUMP: 126 blocks (63KB) on 1 volume at 1000 KB/sec
DUMP: DUMP IS DONE
DUMP: Level 0 dump on 2001년 4월 30일 월요일 오전 11시 45분 27초

예) 2번째 volume의 내용에 backup 을 수행하는 법
e250 [ root{76} /home/zeronine ] mt -f /dev/rmt/0 rew
e250 [ root{76} /home/zeronine ] mt -f /dev/rmt/0n fsf 1
e250 [ root{77} /home/zeronine ] ufsdump 0uf /dev/rmt/0n ./comtec
DUMP: Writing 32 Kilobyte records
DUMP: Date of this level 0 dump: 2001년 5월 2일 수요일 오후 06시 40분 30초
DUMP: Date of last level 0 dump: the epoch
DUMP: Dumping /dev/rdsk/c0t0d0s3 (e250:/home) to /dev/rmt/0n.
DUMP: Mapping (Pass I) [regular files]
DUMP: Mapping (Pass II) [directories]
DUMP: Estimated 1756 blocks (878KB).
DUMP: Dumping (Pass III) [directories]
DUMP: Dumping (Pass IV) [regular files]
DUMP: 1726 blocks (863KB) on 1 volume at 2458 KB/sec
DUMP: DUMP IS DONE

- Restore
ufsrestore의 사용법은 아래와 같다.
ufsrestore [option] [source dump file] [volume number]

option
i Perform an interactive restore(하나하나씩 물어가면서하는 것)
r Restore the entire backup 전체다 백업파일을 restore
t List the table of contents of the backup.
x Restore only the files named on the command line.

a archive_file
Take the table of contents information from the named archive_file rather than the tape. Until you actually need to extract a file, the backup volume does not need to be mounted.
f dump_file
Specify a specific dump_file instead of the default /dev/rmt/0 as the file to restore from. Typically, the dump_file option specifies a tape or diskette drive, but it may be a disk file.

s n Skip to the dump_file indicated by the number and use it for the restoration.

v Display path names as they are being restored(verbose mode). Note that this sis different from the v(verify) option used with the ufsdump command.

restore시에는 명령을 수행한 곳을 기준으로(상대경로로) restore하기 때문에, restore할 때 명령을 수행할 장소를 잘 선택해야 한다.

예) restore하기 전에 tape에 backup된 내용을 확인하고 restore하는 방법.
e250 [ zeronine{3} ~ ] ufsrestore tf /dev/rmt/0
2 .
3 ./lost+found
4 ./min
5 ./nice
6 ./good
7 ./zero
8 ./test
e250 [ root{48} / ] mkdir test_fs_back
e250 [ root{49} / ] cd test_fs_back/
e250 [ root{50} /test_fs_back ] ls
총 0
e250 [ root{51} /test_fs_back ] pwd
/test_fs_back

e250 [ root{52} /test_fs_back ] ufsrestore xf /dev/rmt/0
You have not read any volumes yet.
Unless you know which volume your file(s) are on you should start
with the last volume and work towards the first.
Specify next volume #: 1 <-볼륨은 "1"부터 시작이다.
set owner/mode for '.'? [yn] y
e250 [ root{53} /test_fs_back ] ls -l
총 4
0 -rw-r--r-- 1 root other 0 3월 27일 08:23 good
2 drwx------ 2 root root 512 3월 27일 08:12 lost+found/
0 -rw-r--r-- 1 root other 0 3월 27일 08:22 min
0 -rw-r--r-- 1 root other 0 3월 27일 08:22 nice
2 -rw-r--r-- 1 root other 17 4월 30일 11:44 test
0 -rw-r--r-- 1 root other 0 3월 27일 08:23 zero

예 ) interactive하게 backup자료를 복구하는 방법
e250 [ root{82} /test_fs_back ] ls
총 0
e250 [ root{83} /test_fs_back ] ufsrestore ivf /dev/rmt/0
Verify volume and initialize maps
Media block size is 64
Dump date: Mon Apr 30 11:45:27 2001
Dumped from: the epoch
Level 0 dump of /test_fs on e250:/dev/dsk/c0t8d0s0
Label: none
Extract directories from tape
Initialize symbol table.
ufsrestore > ls
.:
2 *./ 6 good 4 min 8 test
2 *../ 3 lost+found/ 5 nice 7 zero
ufsrestore > add test zero min
ufsrestore > ls
.:
2 *./ 6 good 4 *min 8 *test
2 *../ 3 lost+found/ 5 nice 7 *zero
ufsrestore > delete min
ufsrestore > ls
.:
2 *./ 6 good 4 min 8 *test
2 *../ 3 lost+found/ 5 nice 7 *zero
ufsrestore > extract
Extract requested files
You have not read any volumes yet.
Unless you know which volume your file(s) are on you should start
with the last volume and work towards the first.
Specify next volume #: 1
extract file ./zero
extract file ./test
Add links
Set directory mode, owner, and times.
set owner/mode for '.'? [yn] n
ufsrestore > quit
e250 [ root{84} /test_fs_back ] ls
총 2 2 test 0 zero

예) 두번째 volume에 있는 내용을 restore하는 방법
e250 [ root{79} /home/zeronine ] mt -f /dev/rmt/0 rew
e250 [ root{80} /home/zeronine ] ufsrestore tsf /dev/rmt/0 1
Dump number must be a positive integer
e250 [ root{81} /home/zeronine ] ufsrestore tfs /dev/rmt/0 1
2 .
3 ./lost+found
4 ./min
5 ./nice
6 ./good
7 ./zero
8 ./test
e250 [ root{82} /home/zeronine ] ufsrestore tfs /dev/rmt/0 2
2 .
33033 ./zeronine
247697 ./zeronine/comtec
247698 ./zeronine/comtec/messages
247699 ./zeronine/comtec/vxprint_ht.result
247700 ./zeronine/comtec/vxprint_ht_datadg.result
247701 ./zeronine/comtec/tst_dsk.result
247702 ./zeronine/comtec/tst_dsk_r.res
e250 [ root{90} /comtec_back ] pwd
/comtec_back
e250 [ root{91} /comtec_back ] ufsrestore xfs /dev/rmt/0 2
You have not read any volumes yet.
Unless you know which volume your file(s) are on you should start
with the last volume and work towards the first.
Specify next volume #: 1
set owner/mode for '.'? [yn] n
e250 [ root{92} /comtec_back ] ls
총 2 2 zeronine/
e250 [ root{93} /comtec_back ] cd zeronine
e250 [ root{94} /comtec_back/zeronine ] ls
총 2 2 comtec/
e250 [ root{95} /comtec_back/zeronine ] cd comtec
e250 [ root{96} /comtec_back/zeronine/comtec ] ls
총 1522 240 tst_dsk_r.res
1248 messages 8 vxprint_ht.result
2 tst_dsk.result 24 vxprint_ht_datadg.result

How to Restore the root(/) File System

먼저 ufsdump를 사용하여 / file system을 백업 받아 놓고 시작한다.
1. Load and boot the release media to run level S.
ok boot cdrom -s

2. Create the new file system, if necessary.

3. Check the file system with fsck

4. Mount the file system to /a and change to that directory
# mount /dev/dsk/c0t3d0s0 /a
# cd /a

5. Restore the root(/) file system.
# ufsrestore rvf /dev/rmt/0

6. Remove the restoresymtable.

7. Run the installboot program to create the boot block.
# cd /usr/platform/'uname -i'/lib/fs/ufs
# installboot bootblk /dev/rdsk/c0t3d0s0

8. Unmount the new file system.
# cd /
# umount /a

9. Check the restored file system with fsck.
# fsck /dev/rdsk/c0t3d0s0

10. Reboot the system.

목록 | 입력 | 수정 | 답변 | 삭제