203.2. Maintaining a Linux filesystem

203.2 Maintaining a Linux filesystem​
Weight:3
Description: Candidates should be able to properly maintain a Linux filesystem using system utilities. This objective includes manipulating standard filesystems and monitoring SMART devices.
Key Knowledge Areas:
Tools and utilities to manipulate and ext2, ext3 and ext4
Tools and utilities to perform basic Btrfs operations, including subvolumes and snapshots
Tools and utilities to manipulate XFS
Awareness of ZFS
Terms and Utilities:
mkfs (mkfs.*)
mkswap
fsck (fsck.*)
tune2fs, dumpe2fs and debugfs
btrfs, btrfs-convert
xfs_info, xfs_check, xfs_repair, xfsdump and xfsrestore
smartd, smartctl
File Systems Overview
Since lpic 1 course we have got familiar with some File Systems. Lets Compare important ones:
File System
Architecture
Max File Size
Max Volume Size
extends
journal
snapshot
ext2 (1993)
h-tree
2 Tbytes
32 Tbytes
No
No
No
ext3 (2001)
h-tree
2 Tbytes
32 Tbytes
No
Yes
No
ext4 (2006)
h-tree
16 Tbytes
1 Ebytes
Yes
Yes
No
xfs (1994)
b-tree
8 Ebytes
8 Ebytes
No
Yes
Planned
btrfs (2009)
b-tree
16 Ebytes
16 Ebytes
Yes
Yes
Yes
mkfs
mkfs (Make a file System) command is used to create a File System on a device.To review LPIC1:
1
[email protected]:~# lsblk
2
NAME   MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
3
sdb      8:16   0   10G  0 disk 
4
└─sdb1   8:17   0   10G  0 part 
5
sr0     11:0    1 1024M  0 rom  
6
sda      8:0    0   50G  0 disk 
7
├─sda2   8:2    0    1K  0 part 
8
├─sda5   8:5    0 1021M  0 part [SWAP]
9
└─sda1   8:1    0   49G  0 part /
10
​
11
[email protected]:~# which mkfs
12
/sbin/mkfs
Copied!
We saw that mkfs is a front end for other commands:
1
[email protected]:~# ls /sbin | grep mkfs
2
mkfs
3
mkfs.bfs
4
mkfs.cramfs
5
mkfs.ext2
6
mkfs.ext3
7
mkfs.ext4
8
mkfs.ext4dev
9
mkfs.fat
10
mkfs.minix
11
mkfs.msdos
12
mkfs.ntfs
13
mkfs.vfat
Copied!
We can use mkfs.ext4 /dev/sdb1 to format a partition or in another format:
1
[email protected]:~# mkfs -t ext4 /dev/sdb1
2
mke2fs 1.42.13 (17-May-2015)
3
Creating filesystem with 2621184 4k blocks and 655360 inodes
4
Filesystem UUID: 3f940f9b-d46d-423e-9f20-4cf74564e7ac
5
Superblock backups stored on blocks: 
6
    32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632
7
​
8
Allocating group tables: done                            
9
Writing inode tables: done                            
10
Creating journal (32768 blocks): done
11
Writing superblocks and filesystem accounting information: done
Copied!
mkswap
There is an exception among File Systems which is swap. We have mentioned that swap can be a file or an entire partition. To make a swap partition use mkswap /dev/sdXY command
File System maintenance Tools
Each file system has it own tools to maintain the problems.
fsck
fsck ( File System consistency check) is a tool for checking and repairing linux file system. Like mkfs, fsck is a front of other commands.
1
[email protected]:~# which fsck
2
/sbin/fsck
3
[email protected]:~# ls /sbin | grep fsck
4
dosfsck
5
e2fsck
6
fsck
7
fsck.cramfs
8
fsck.ext2
9
fsck.ext3
10
fsck.ext4
11
fsck.ext4dev
12
fsck.fat
13
fsck.minix
14
fsck.msdos
15
fsck.nfs
16
fsck.vfat
17
fsck.xfs
Copied!
fsck is originally an ext tool . Try not to use fsck for other file systems because it might cause damage or data lost! How ever it is smart about other file systems like XFS and avoid running on them.
1
[email protected]:~# mount /dev/sdb1 /mnt/my10gdisk/
2
[email protected]:~# lsblk
3
NAME   MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
4
sdb      8:16   0   10G  0 disk 
5
└─sdb1   8:17   0   10G  0 part /mnt/my10gdisk
6
sr0     11:0    1 1024M  0 rom  
7
sda      8:0    0   50G  0 disk 
8
├─sda2   8:2    0    1K  0 part 
9
├─sda5   8:5    0 1021M  0 part [SWAP]
10
└─sda1   8:1    0   49G  0 part /
11
[email protected]:~# fsck /dev/sdb1
12
fsck from util-linux 2.27.1
13
e2fsck 1.42.13 (17-May-2015)
14
/dev/sdb1 is mounted.
15
e2fsck: Cannot continue, aborting.
Copied!
Do not forget you can not check a filesystem while you are using that, so do not forget to unmount!!!
1
[email protected]:~# umount /mnt/my10gdisk 
2
[email protected]:~# fsck /dev/sdb1
3
fsck from util-linux 2.27.1
4
e2fsck 1.42.13 (17-May-2015)
5
/dev/sdb1: clean, 11/655360 files, 79663/2621184 blocks
Copied!
fsck exit codes:
Based on health of ext Partition, different exit codes might be generated by fsck:
1
0-No errors
2
1-Filesystem errors corrected
3
2-System should be rebooted
4
4-Filesystem errors left uncorrected
5
8-Operational error
6
16-Usage or syntax error
7
32-Checking canceled by user request
8
128-Shared-library error
Copied!
fsck usefull switeches
Description
-v
Produce Verbose output, include all file-system specific commands that are executed
-f
force fixing errors, ask conformation before each repair
-y
say "yes" to confirm fixing all errors
-n
emulate fixing errors, but no real write on hard disk
-N
Tell what fsck command will be run, and nothing is exexuted
-b <SuperBlockBackupLocation>
Restore super block from the backup
tune2fs
tune2fs Change different aspects of file system
1
[email protected]:~# tune2fs --help
2
tune2fs 1.42.13 (17-May-2015)
3
tune2fs: invalid option -- '-'
4
Usage: tune2fs [-c max_mounts_count] [-e errors_behavior] [-g group]
5
    [-i interval[d|m|w]] [-j] [-J journal_options] [-l]
6
    [-m reserved_blocks_percent] [-o [^]mount_options[,...]] [-p mmp_update_interval]
7
    [-r reserved_blocks_count] [-u user] [-C mount_count] [-L volume_label]
8
    [-M last_mounted_dir] [-O [^]feature[,...]]
9
    [-Q quota_options]
10
    [-E extended-option[,...]] [-T last_check_time] [-U UUID]
11
    [ -I new_inode_size ] device
Copied!
fsck , tune2fs , fstab relation
Previously we see the fstab file for making automatic mounts, in the last field of fstab file we saw pass parameter:
1
# <file system> <mount point>   <type>  <options>       <dump>  <pass>
2
# / was on /dev/sda1 during installation
3
UUID=142a64e5-96f3-4789-9c91-1dc1570057b7 /               ext4    errors=remount-ro 0       1
Copied!
We discussed then that the pass parameter can be 0, 1 or 2.
0
Never be checked by fsck during boot process
1
Check the partition with fsck duting boot with priority 1
2
Check the partition after other partitions with fsck priority2
But if you put 1 in fstab it would not work! Why? Because fsck look at other parameters of Partition File System and check intervals:
1
[email protected]:~# tune2fs -l /dev/sdb1
2
tune2fs 1.42.13 (17-May-2015)
3
Filesystem volume name:   <none>
4
Last mounted on:          <not available>
5
Filesystem UUID:          395e0905-bcc7-4b8d-b3ca-3193c266b749
6
Filesystem magic number:  0xEF53
7
Filesystem revision #:    1 (dynamic)
8
Filesystem features:      has_journal ext_attr resize_inode dir_index filetype extent flex_bg sparse_super large_file huge_file uninit_bg dir_nlink extra_isize
9
Filesystem flags:         signed_directory_hash 
10
Default mount options:    user_xattr acl
11
Filesystem state:         clean
12
Errors behavior:          Continue
13
Filesystem OS type:       Linux
14
Inode count:              655360
15
Block count:              2621184
16
Reserved block count:     131059
17
Free blocks:              2541521
18
Free inodes:              655349
19
First block:              0
20
Block size:               4096
21
Fragment size:            4096
22
Reserved GDT blocks:      639
23
Blocks per group:         32768
24
Fragments per group:      32768
25
Inodes per group:         8192
26
Inode blocks per group:   512
27
Flex block group size:    16
28
Filesystem created:       Sat Dec 30 04:03:38 2017
29
Last mount time:          Sat Dec 30 04:03:54 2017
30
Last write time:          Sat Dec 30 04:44:35 2017
31
Mount count:              0
32
Maximum mount count:      -1
33
Last checked:             Sat Dec 30 04:44:35 2017
34
Check interval:           0 (<none>)
35
Lifetime writes:          291 MB
36
Reserved blocks uid:      0 (user root)
37
Reserved blocks gid:      0 (group root)
38
First inode:              11
39
Inode size:              256
40
Required extra isize:     28
41
Desired extra isize:      28
42
Journal inode:            8
43
Default directory hash:   half_md4
44
Directory Hash Seed:      54a34eb1-0de8-40af-8e37-e4817b1f3046
45
Journal backup:           inode blocks
Copied!
Among this long list Check interval and mount count are realy important. fsck check partitions we have mentioned in fstab if specific time interval arrives or number of mount count reach.
usefull commands
Description
tune2fs -L  KING8GIG /dev/sdc1
Change Volume Name
tune2fs -i 7 /dev/sdb1
Change check interval to 7 days
tune2fs -c 10 /dev/sdb1
Change mount count to 10 times
tune2fs -m 0 /dev/sdb1
Change  reserved Blocks to 0
dumpe2fs
dumpe2fs Show all super blocks info. SuperBlocks are kind of meta data of File System which keeps information about size, block size, the empty and filled blocks, size and location of inodes and many many more information. dumpe2fs can be used while partition is mounted.
1
[email protected]:~# dumpe2fs /dev/sdb1
2
dumpe2fs 1.42.13 (17-May-2015)
3
Filesystem volume name:   <none>
4
Last mounted on:          <not available>
5
Filesystem UUID:          395e0905-bcc7-4b8d-b3ca-3193c266b749
6
Filesystem magic number:  0xEF53
7
Filesystem revision #:    1 (dynamic)
8
Filesystem features:      has_journal ext_attr resize_inode dir_index filetype extent flex_bg sparse_super large_file huge_file uninit_bg dir_nlink extra_isize
9
Filesystem flags:         signed_directory_hash 
10
Default mount options:    user_xattr acl
11
Filesystem state:         clean
12
Errors behavior:          Continue
13
Filesystem OS type:       Linux
14
Inode count:              655360
15
Block count:              2621184
16
Reserved block count:     131059
17
Free blocks:              2541521
18
Free inodes:              655349
19
First block:              0
20
Block size:               4096
21
Fragment size:            4096
22
Reserved GDT blocks:      639
23
Blocks per group:         32768
24
Fragments per group:      32768
25
Inodes per group:         8192
26
Inode blocks per group:   512
27
Flex block group size:    16
28
Filesystem created:       Sat Dec 30 04:03:38 2017
29
Last mount time:          Sat Dec 30 04:03:54 2017
30
Last write time:          Sat Dec 30 04:05:42 2017
31
Mount count:              1
32
Maximum mount count:      -1
33
Last checked:             Sat Dec 30 04:03:38 2017
34
Check interval:           0 (<none>)
35
Lifetime writes:          291 MB
36
Reserved blocks uid:      0 (user root)
37
Reserved blocks gid:      0 (group root)
38
First inode:              11
39
Inode size:              256
40
Required extra isize:     28
41
Desired extra isize:      28
42
Journal inode:            8
43
Default directory hash:   half_md4
44
Directory Hash Seed:      54a34eb1-0de8-40af-8e37-e4817b1f3046
45
Journal backup:           inode blocks
46
Journal features:         (none)
47
Journal size:             128M
48
Journal length:           32768
49
Journal sequence:         0x00000004
50
Journal start:            0
51
​
52
​
53
Group 0: (Blocks 0-32767) [ITABLE_ZEROED]
54
  Checksum 0xe944, unused inodes 8181
55
  Primary superblock at 0, Group descriptors at 1-1
56
  Reserved GDT blocks at 2-640
57
  Block bitmap at 641 (+641), Inode bitmap at 657 (+657)
58
  Inode table at 673-1184 (+673)
59
  23897 free blocks, 8181 free inodes, 2 directories, 8181 unused inodes
60
  Free blocks: 8871-32767
61
  Free inodes: 12-8192
62
Group 1: (Blocks 32768-65535) [INODE_UNINIT, BLOCK_UNINIT, ITABLE_ZEROED]
63
  Checksum 0xda21, unused inodes 8192
64
  Backup superblock at 32768, Group descriptors at 32769-32769
65
  Reserved GDT blocks at 32770-33408
66
  Block bitmap at 642 (bg #0 + 642), Inode bitmap at 658 (bg #0 + 658)
67
  Inode table at 1185-1696 (bg #0 + 1185)
68
  32127 free blocks, 8192 free inodes, 0 directories, 8192 unused inodes
69
  Free blocks: 33409-65535
70
  Free inodes: 8193-16384
71
Group 2: (Blocks 65536-98303) [INODE_UNINIT, BLOCK_UNINIT, ITABLE_ZEROED]
72
  Checksum 0x8c02, unused inodes 8192
73
  Block bitmap at 643 (bg #0 + 643), Inode bitmap at 659 (bg #0 + 659)
74
  Inode table at 1697-2208 (bg #0 + 1697)
75
  32768 free blocks, 8192 free inodes, 0 directories, 8192 unused inodes
76
  Free blocks: 65536-98303
77
  Free inodes: 16385-24576
78
.
79
..
80
...
81
​
82
Group 78: (Blocks 2555904-2588671) [INODE_UNINIT, BLOCK_UNINIT, ITABLE_ZEROED]
83
  Checksum 0x2104, unused inodes 8192
84
  Block bitmap at 2097166 (bg #64 + 14), Inode bitmap at 2097182 (bg #64 + 30)
85
  Inode table at 2104352-2104863 (bg #64 + 7200)
86
  32768 free blocks, 8192 free inodes, 0 directories, 8192 unused inodes
87
  Free blocks: 2555904-2588671
88
  Free inodes: 638977-647168
89
Group 79: (Blocks 2588672-2621183) [INODE_UNINIT, ITABLE_ZEROED]
90
  Checksum 0x8657, unused inodes 8192
91
  Block bitmap at 2097167 (bg #64 + 15), Inode bitmap at 2097183 (bg #64 + 31)
92
  Inode table at 2104864-2105375 (bg #64 + 7712)
93
  32512 free blocks, 8192 free inodes, 0 directories, 8192 unused inodes
94
  Free blocks: 2588672-2621183
95
  Free inodes: 647169-655360
Copied!
As superbloacks are so important Files System is designed to keep back of them by copying them on several places of hard disk.
1
[email protected]:~# dumpe2fs /dev/sdb1 | grep -i Superblock
2
dumpe2fs 1.42.13 (17-May-2015)
3
  Primary superblock at 0, Group descriptors at 1-1
4
  Backup superblock at 32768, Group descriptors at 32769-32769
5
  Backup superblock at 98304, Group descriptors at 98305-98305
6
  Backup superblock at 163840, Group descriptors at 163841-163841
7
  Backup superblock at 229376, Group descriptors at 229377-229377
8
  Backup superblock at 294912, Group descriptors at 294913-294913
9
  Backup superblock at 819200, Group descriptors at 819201-819201
10
  Backup superblock at 884736, Group descriptors at 884737-884737
11
  Backup superblock at 1605632, Group descriptors at 1605633-1605633
Copied!
There is another way to observe location of super blocks backup:
1
[email protected]:~# mkfs.ext4 -n /dev/sdb1
2
mke2fs 1.42.13 (17-May-2015)
3
/dev/sdb1 contains a ext4 file system
4
    last mounted on Sat Dec 30 04:03:54 2017
5
Proceed anyway? (y,n) y
6
Creating filesystem with 2621184 4k blocks and 655360 inodes
7
Filesystem UUID: 9979d6a3-985e-4435-b445-f399f3715e79
8
Superblock backups stored on blocks: 
9
    32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632
Copied!
-n cuase no changes on Hard disk and just show what it would do if it were to create a file system. And finally try dumpe2fs -h to get information about super blocks. Try it your self :)
debugfs
debugfs interactive file system editor.Its better to unmount the partition first.
1
[email protected]:~# debugfs /dev/sdb1
2
debugfs 1.42.13 (17-May-2015)
3
debugfs:  help
4
​
5
Available debugfs requests:
6
​
7
show_debugfs_params, params
8
                         Show debugfs parameters
9
open_filesys, open       Open a filesystem
10
close_filesys, close     Close the filesystem
11
freefrag, e2freefrag     Report free space fragmentation
12
feature, features        Set/print superblock features
13
dirty_filesys, dirty     Mark the filesystem as dirty
14
init_filesys             Initialize a filesystem (DESTROYS DATA)
15
show_super_stats, stats  Show superblock statistics
16
ncheck                   Do inode->name translation
17
icheck                   Do block->inode translation
18
change_root_directory, chroot
19
                         Change root directory
20
change_working_directory, cd
21
                         Change working directory
22
list_directory, ls       List directory
23
show_inode_info, stat    Show inode information 
24
dump_extents, extents, ex
25
                         Dump extents information 
26
blocks                   Dump blocks used by an inode 
27
filefrag                 Report fragmentation information for an inode
28
link, ln                 Create directory link
29
unlink                   Delete a directory link
30
mkdir                    Create a directory
31
rmdir                    Remove a directory
32
rm                       Remove a file (unlink and kill_file, if appropriate)
33
kill_file                Deallocate an inode and its blocks
34
clri                     Clear an inode's contents
35
freei                    Clear an inode's in-use flag
36
seti                     Set an inode's in-use flag
37
testi                    Test an inode's in-use flag
38
freeb                    Clear a block's in-use flag
39
setb                     Set a block's in-use flag
40
testb                    Test a block's in-use flag
41
modify_inode, mi         Modify an inode by structure
42
find_free_block, ffb     Find free block(s)
43
find_free_inode, ffi     Find free inode(s)
44
print_working_directory, pwd
45
                         Print current working directory
46
expand_dir, expand       Expand directory
47
mknod                    Create a special file
48
list_deleted_inodes, lsdel
49
                         List deleted inodes
50
undelete, undel          Undelete file
51
write                    Copy a file from your native filesystem
52
dump_inode, dump         Dump an inode out to a file
53
cat                      Dump an inode out to stdout
54
lcd                      Change the current directory on your native filesystem
55
rdump                    Recursively dump a directory to the native filesystem
56
set_super_value, ssv     Set superblock value
57
set_inode_field, sif     Set inode field
58
set_block_group, set_bg  Set block group descriptor field
59
logdump                  Dump the contents of the journal
60
htree_dump, htree        Dump a hash-indexed directory
61
dx_hash, hash            Calculate the directory hash of a filename
62
dirsearch                Search a directory for a particular filename
63
bmap                     Calculate the logical->physical block mapping for an inode
64
punch, truncate          Punch (or truncate) blocks from an inode by deallocating them
65
symlink                  Create a symbolic link
66
imap                     Calculate the location of an inode
67
dump_unused              Dump unused blocks
68
set_current_time         Set current time to use when setting filesystem fields
69
supported_features       Print features supported by this version of e2fsprogs
70
dump_mmp                 Dump MMP information
71
set_mmp_value, smmp      Set MMP value
72
extent_open, eo          Open inode for extent manipulation
73
zap_block, zap           Zap block: fill with 0, pattern, flip bits etc.
74
block_dump, bdump, bd    Dump contents of a block
75
list_quota, lq           List quota
76
get_quota, gq            Get quota
77
inode_dump, idump, id    Dump the inode structure in hex
78
help                     Display info on command or topic.
79
list_requests, lr, ?     List available commands.
80
quit, q                  Leave the subsystem.
Copied!
maintenance XFS File System
XFS File System has its own tools.Base on our distro we might need to install xfsprogs to use XFS tools but XFS is Redhat Favorite File system so we use centOS7 here:
1
[[email protected] ~]# lsblk
2
NAME            MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
3
sda               8:0    0   50G  0 disk 
4
├─sda1            8:1    0    1G  0 part /boot
5
└─sda2            8:2    0   49G  0 part 
6
  ├─centos-root 253:0    0 45.1G  0 lvm  /
7
  └─centos-swap 253:1    0  3.9G  0 lvm  [SWAP]
8
sdb               8:16   0   10G  0 disk 
9
└─sdb1            8:17   0   10G  0 part 
10
sr0              11:0    1 1024M  0 rom  
11
[[email protected] ~]# mkfs.xfs /dev/sdb1
12
meta-data=/dev/sdb1              isize=512    agcount=4, agsize=655296 blks
13
         =                       sectsz=512   attr=2, projid32bit=1
14
         =                       crc=1        finobt=0, sparse=0
15
data     =                       bsize=4096   blocks=2621184, imaxpct=25
16
         =                       sunit=0      swidth=0 blks
17
naming   =version 2              bsize=4096   ascii-ci=0 ftype=1
18
log      =internal log           bsize=4096   blocks=2560, version=2
19
         =                       sectsz=512   sunit=0 blks, lazy-count=1
20
realtime =none                   extsz=4096   blocks=0, rtextents=0
21
​
22
[[email protected] ~]# xfs
23
xfs_admin      xfs_estimate   xfsinvutil     xfs_mkfile     xfs_rtcp
24
xfs_bmap       xfs_freeze     xfs_io         xfs_ncheck     
25
xfs_copy       xfs_fsr        xfs_logprint   xfs_quota      
26
xfs_db         xfs_growfs     xfs_mdrestore  xfs_repair     
27
xfsdump        xfs_info       xfs_metadump   xfsrestore
Copied!
Okey we have Formatted and re partitioned /dev/sdb1 lets get to know about some XFS File System tools:
xfsprogs
Description
xfs_info
Give info about XFS Partition, partition must be mounted
xfs_repair
Fix XFS File System problem
xfs_check
Depricated ! Check XFS File System problems with no verbosity, use -v or "xfs_repair -n" instead
xfsdump & xfsrestore
XFS backup and restore commands, They can work with mounted partition as they are Backup & Restore process
xfs_info
1
[[email protected] ~]# mkdir /mnt/my10gxfs
2
[[email protected] ~]# mount /dev/sdb1 /mnt/my10gxfs/
3
[[email protected] ~]# xfs_info /mnt/my10gxfs/
4
meta-data=/dev/sdb1              isize=512    agcount=4, agsize=655296 blks
5
         =                       sectsz=512   attr=2, projid32bit=1
6
         =                       crc=1        finobt=0 spinodes=0
7
data     =                       bsize=4096   blocks=2621184, imaxpct=25
8
         =                       sunit=0      swidth=0 blks
9
naming   =version 2              bsize=4096   ascii-ci=0 ftype=1
10
log      =internal               bsize=4096   blocks=2560, version=2
11
         =                       sectsz=512   sunit=0 blks, lazy-count=1
12
realtime =none                   extsz=4096   blocks=0, rtextents=0
Copied!
xfs_repair
1
[[email protected] ~]# umount /mnt/my10gxfs/
2
​
3
[[email protected] ~]# xfs_repair -n /dev/sdb1
4
Phase 1 - find and verify superblock...
5
Phase 2 - using internal log
6
        - zero log...
7
        - scan filesystem freespace and inode maps...
8
        - found root inode chunk
9
Phase 3 - for each AG...
10
        - scan (but don't clear) agi unlinked lists...
11
        - process known inodes and perform inode discovery...
12
        - agno = 0
13
        - agno = 1
14
        - agno = 2
15
        - agno = 3
16
        - process newly discovered inodes...
17
Phase 4 - check for duplicate blocks...
18
        - setting up duplicate extent list...
19
        - check for inodes claiming duplicate blocks...
20
        - agno = 0
21
        - agno = 2
22
        - agno = 3
23
        - agno = 1
24
No modify flag set, skipping phase 5
25
Phase 6 - check inode connectivity...
26
        - traversing filesystem ...
27
        - traversal finished ...
28
        - moving disconnected inodes to lost+found ...
29
Phase 7 - verify link counts...
30
No modify flag set, skipping filesystem flush and exiting.
Copied!
and if any problems have found use xfs_repair for that.
xfsdump , xfsrestore
There are two old xfs commands for backing up and restoring. They history back to days when tape media were only backup media.
1
[[email protected] ~]# lsblk
2
NAME            MAJ:MIN RM  SIZE RO TYPE MOUNTPOINT
3
sda               8:0    0   50G  0 disk 
4
├─sda1            8:1    0    1G  0 part /boot
5
└─sda2            8:2    0   49G  0 part 
6
  ├─centos-root 253:0    0 45.1G  0 lvm  /
7
  └─centos-swap 253:1    0  3.9G  0 lvm  [SWAP]
8
sdb               8:16   0   10G  0 disk 
9
└─sdb1            8:17   0   10G  0 part 
10
sr0              11:0    1 1024M  0 rom  
11
[[email protected] ~]# mount -t xfs /dev/sdb1 /mnt/my10gxfs/
12
[[email protected] ~]# mkdir /mnt/mybackups
13
[[email protected] ~]# cd /mnt/mybackups/
14
​
15
[[email protected] mybackups]# xfsdump -f /mnt/mybackups/backup1jun18 /mnt/my10gxfs
16
xfsdump: using file dump (drive_simple) strategy
17
xfsdump: version 3.1.4 (dump format 3.0) - type ^C for status and control
18
​
19
 ============================= dump label dialog ==============================
20
​
21
please enter label for this dump session (timeout in 300 sec)
22
 -> etcbackup
23
session label entered: "etcbackup"
24
​
25
 --------------------------------- end dialog ---------------------------------
26
​
27
xfsdump: level 0 dump of server1:/mnt/my10gxfs
28
xfsdump: dump date: Mon Jan  1 02:03:26 2018
29
xfsdump: session id: 482cd1f2-eed5-492c-bf8b-d3befb97be97
30
xfsdump: session label: "etcbackup"
31
xfsdump: ino map phase 1: constructing initial dump list
32
xfsdump: ino map phase 2: skipping (no pruning necessary)
33
xfsdump: ino map phase 3: skipping (only one dump stream)
34
xfsdump: ino map construction complete
35
xfsdump: estimated dump size: 20800 bytes
36
xfsdump: /var/lib/xfsdump/inventory created
37
​
38
 ============================= media label dialog =============================
39
​
40
please enter label for media in drive 0 (timeout in 300 sec)
41
​
42
-> bak1jun18label
43
media label entered: "bak1jun18label"
44
​
45
 --------------------------------- end dialog ---------------------------------
46
​
47
xfsdump: creating dump session media file 0 (media 0, file 0)
48
xfsdump: dumping ino map
49
xfsdump: dumping directories
50
xfsdump: dumping non-directory files
51
xfsdump: ending media file
52
xfsdump: media file size 21016 bytes
53
xfsdump: dump size (non-dir files) : 0 bytes
54
xfsdump: dump complete: 201 seconds elapsed
55
xfsdump: Dump Summary:
56
xfsdump:   stream 0 /mnt/mybackups/backup1jun18 OK (success)
57
xfsdump: Dump Status: SUCCESS
Copied!
There Three levels of backups that can be done woth xfsdump
xfsdump backup level
descriptio
0
Full backup
1
Only backup changed Files since Last backup
2
Backup File Changes since last full or Incremental Backup
Lets see the result:
1
[[email protected] mybackups]# ls
2
backup1jun18
3
[[email protected] mybackups]# cat backup1jun18 
4
xFSdump0"v��ZI�>������zH,����I,��Ӿ����server1etcbackupQbak1jun18label�z]4��K7�?`%�\s�/mnt/my10gxfs/dev/sdb1xfs�n$��tG��P���ᳵ?�@@@#��[email protected]�ZI�ÎZI�Î��������
Copied!
and try to restore:
1
[[email protected] mnt]# xfsrestore -f /mnt/mybackups/backup1jun18 /mnt/restore/
2
xfsrestore: using file dump (drive_simple) strategy
3
xfsrestore: version 3.1.4 (dump format 3.0) - type ^C for status and control
4
xfsrestore: searching media for dump
5
xfsrestore: examining media file 0
6
xfsrestore: dump description: 
7
xfsrestore: hostname: server1
8
xfsrestore: mount point: /mnt/my10gxfs
9
xfsrestore: volume: /dev/sdb1
10
xfsrestore: session time: Mon Jan  1 02:03:26 2018
11
xfsrestore: level: 0
12
xfsrestore: session label: "etcbackup"
13
xfsrestore: media label: "bak1jun18label"
14
xfsrestore: file system id: 966e24bf-dc74-47ba-a250-80fe84e1b3b5
15
xfsrestore: session id: 482cd1f2-eed5-492c-bf8b-d3befb97be97
16
xfsrestore: media id: e87a5d34-a49b-4b37-a33f-6025ab5c73eb
17
xfsrestore: using online session inventory
18
xfsrestore: searching media for directory dump
19
xfsrestore: reading directories
20
xfsrestore: 1 directories and 0 entries processed
21
xfsrestore: directory post-processing
22
xfsrestore: restore complete: 0 seconds elapsed
23
xfsrestore: Restore Summary:
24
xfsrestore:   stream 0 /mnt/mybackups/backup1jun18 OK (success)
25
xfsrestore: Restore Status: SUCCESS
26
​
27
[[email protected] mnt]# cd my10gxfs/
28
[[email protected] my10gxfs]# ls
Copied!
As you might not be an only admin of system use xfsrestore -I command to have a list of what ever has been backed up with xfsdump, including file sessions IDs and backup session IDs:
1
[[email protected] my10gxfs]# xfsrestore -I
2
file system 0:
3
    fs id:        966e24bf-dc74-47ba-a250-80fe84e1b3b5
4
    session 0:
5
        mount point:    server1:/mnt/my10gxfs
6
        device:        server1:/dev/sdb1
7
        time:        Mon Jan  1 02:03:26 2018
8
        session label:    "etcbackup"
9
        session id:    482cd1f2-eed5-492c-bf8b-d3befb97be97
10
        level:        0
11
        resumed:    NO
12
        subtree:    NO
13
        streams:    1
14
        stream 0:
15
            pathname:    /mnt/mybackups/backup1jun18
16
            start:        ino 0 offset 0
17
            end:        ino 1 offset 0
18
            interrupted:    NO
19
            media files:    1
20
            media file 0:
21
                mfile index:    0
22
                mfile type:    data
23
                mfile size:    21016
24
                mfile start:    ino 0 offset 0
25
                mfile end:    ino 1 offset 0
26
                media label:    "bak1jun18label"
27
                media id:    e87a5d34-a49b-4b37-a33f-6025ab5c73eb
28
xfsrestore: Restore Status: SUCCESS
Copied!
smartd , smartctl
S.M.A.R.T. (Self-Monitoring, Analysis and Reporting Technology) is a monitoring system included in computer hard disk drives (HDDs) and even solid-state drives (SSDs).So it is not linux specific tool or feature . it its duty is monitoring various indicators of Drive reliability in order to avoid hardware failures . Usually there is a software or 3rd party utility which is running on Host operating system which reads SMART data and notify user about health status of Disk Drive to avoid data lost.
Linux has two tiny tools smartd and smartctl for reading SMART data.Based on your distro you might need install smartmontool.
1
[email protected]:~# apt install smartmontools
Copied!
and to check weather our hard disk supports SMART and its enabled:
1
[email protected]:~# smartctl -i /dev/sda
2
smartctl 6.5 2016-01-24 r4214 [x86_64-linux-4.10.0-40-generic] (local build)
3
Copyright (C) 2002-16, Bruce Allen, Christian Franke, www.smartmontools.org
4
​
5
=== START OF INFORMATION SECTION ===
6
Model Family:     Samsung based SSDs
7
Device Model:     Samsung SSD 850 EVO 500GB
8
Serial Number:    S2R9NX0H503451W
9
LU WWN Device Id: 5 002538 d40e515dc
10
Firmware Version: EMT02B6Q
11
User Capacity:    500,107,862,016 bytes [500 GB]
12
Sector Size:      512 bytes logical/physical
13
Rotation Rate:    Solid State Device
14
Form Factor:      2.5 inches
15
Device is:        In smartctl database [for details use: -P show]
16
ATA Version is:   ACS-2, ATA8-ACS T13/1699-D revision 4c
17
SATA Version is:  SATA 3.1, 6.0 Gb/s (current: 6.0 Gb/s)
18
Local Time is:    Mon Jan  1 13:37:38 2018 +0330
19
SMART support is: Available - device has SMART capability.
20
SMART support is: Enabled
Copied!
to do the <short|long|...> test on hard disk :
1
[email protected]:~# smartctl -t 
2
afterselect,off  long             scttempint,      select,redo
3
afterselect,on   offline          select,          short
4
conveyance       pending,         select,next      vendor,
5
​
6
[email protected]:~# smartctl -t short /dev/sda
7
smartctl 6.5 2016-01-24 r4214 [x86_64-linux-4.10.0-40-generic] (local build)
8
Copyright (C) 2002-16, Bruce Allen, Christian Franke, www.smartmontools.org
9
​
10
=== START OF OFFLINE IMMEDIATE AND SELF-TEST SECTION ===
11
Sending command: "Execute SMART Short self-test routine immediately in off-line mode".
12
Drive command "Execute SMART Short self-test routine immediately in off-line mode" successful.
13
Testing has begun.
14
Please wait 2 minutes for test to complete.
15
Test will complete after Mon Jan  1 13:45:37 2018
16
​
17
Use smartctl -X to abort test.
Copied!
and for watching results:
1
[email protected]:~# smartctl -a /dev/sda
2
smartctl 6.5 2016-01-24 r4214 [x86_64-linux-4.10.0-40-generic] (local build)
3
Copyright (C) 2002-16, Bruce Allen, Christian Franke, www.smartmontools.org
4
​
5
=== START OF INFORMATION SECTION ===
6
Model Family:     Samsung based SSDs
7
Device Model:     Samsung SSD 850 EVO 500GB
8
Serial Number:    S2R9NX0H503451W
9
LU WWN Device Id: 5 002538 d40e515dc
10
Firmware Version: EMT02B6Q
11
User Capacity:    500,107,862,016 bytes [500 GB]
12
Sector Size:      512 bytes logical/physical
13
Rotation Rate:    Solid State Device
14
Form Factor:      2.5 inches
15
Device is:        In smartctl database [for details use: -P show]
16
ATA Version is:   ACS-2, ATA8-ACS T13/1699-D revision 4c
17
SATA Version is:  SATA 3.1, 6.0 Gb/s (current: 6.0 Gb/s)
18
Local Time is:    Mon Jan  1 14:02:06 2018 +0330
19
SMART support is: Available - device has SMART capability.
20
SMART support is: Enabled
21
​
22
=== START OF READ SMART DATA SECTION ===
23
SMART overall-health self-assessment test result: PASSED
24
​
25
General SMART Values:
26
Offline data collection status:  (0x00)    Offline data collection activity
27
                    was never started.
28
                    Auto Offline Data Collection: Disabled.
29
Self-test execution status:      (   0)    The previous self-test routine completed
30
                    without error or no self-test has ever 
31
                    been run.
32
Total time to complete Offline 
33
data collection:         (    0) seconds.
34
Offline data collection
35
capabilities:              (0x53) SMART execute Offline immediate.
36
                    Auto Offline data collection on/off support.
37
                    Suspend Offline collection upon new
38
                    command.
39
                    No Offline surface scan supported.
40
                    Self-test supported.
41
                    No Conveyance Self-test supported.
42
                    Selective Self-test supported.
43
SMART capabilities:            (0x0003)    Saves SMART data before entering
44
                    power-saving mode.
45
                    Supports SMART auto save timer.
46
Error logging capability:        (0x01)    Error logging supported.
47
                    General Purpose Logging supported.
48
Short self-test routine 
49
recommended polling time:      (   2) minutes.
50
Extended self-test routine
51
recommended polling time:      ( 265) minutes.
52
SCT capabilities:            (0x003d)    SCT Status supported.
53
                    SCT Error Recovery Control supported.
54
                    SCT Feature Control supported.
55
                    SCT Data Table supported.
56
​
57
SMART Attributes Data Structure revision number: 1
58
Vendor Specific SMART Attributes with Thresholds:
59
ID# ATTRIBUTE_NAME          FLAG     VALUE WORST THRESH TYPE      UPDATED  WHEN_FAILED RAW_VALUE
60
  5 Reallocated_Sector_Ct   0x0033   100   100   010    Pre-fail  Always       -       0
61
  9 Power_On_Hours          0x0032   099   099   000    Old_age   Always       -       1136
62
 12 Power_Cycle_Count       0x0032   098   098   000    Old_age   Always       -       1292
63
177 Wear_Leveling_Count     0x0013   099   099   000    Pre-fail  Always       -       6
64
179 Used_Rsvd_Blk_Cnt_Tot   0x0013   100   100   010    Pre-fail  Always       -       0
65
181 Program_Fail_Cnt_Total  0x0032   100   100   010    Old_age   Always       -       0
66
182 Erase_Fail_Count_Total  0x0032   100   100   010    Old_age   Always       -       0
67
183 Runtime_Bad_Block       0x0013   100   100   010    Pre-fail  Always       -       0
68
187 Uncorrectable_Error_Cnt 0x0032   100   100   000    Old_age   Always       -       0
69
190 Airflow_Temperature_Cel 0x0032   068   052   000    Old_age   Always       -       32
70
195 ECC_Error_Rate          0x001a   200   200   000    Old_age   Always       -       0
71
199 CRC_Error_Count         0x003e   100   100   000    Old_age   Always       -       0
72
235 POR_Recovery_Count      0x0012   099   099   000    Old_age   Always       -       72
73
241 Total_LBAs_Written      0x0032   099   099   000    Old_age   Always       -       7397082470
74
​
75
SMART Error Log Version: 1
76
No Errors Logged
77
​
78
SMART Self-test log structure revision number 1
79
Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
80
# 1  Short offline       Completed without error       00%      1135         -
81
# 2  Short offline       Completed without error       00%         1         -
82
​
83
SMART Selective self-test log data structure revision number 1
84
 SPAN  MIN_LBA  MAX_LBA  CURRENT_TEST_STATUS
85
    1        0        0  Not_testing
86
    2        0        0  Not_testing
87
    3        0        0  Not_testing
88
    4        0        0  Not_testing
89
    5        0        0  Not_testing
90
  255        0    65535  Read_scanning was never started
91
Selective self-test flags (0x0):
92
  After scanning selected spans, do NOT read-scan remainder of disk.
93
If Selective self-test is pending on power-up, resume after 0 minute delay.
Copied!
some other usefull smartctl commands:
smartctl command
Description
smartctl -h /dev/sda
Print Help
smartctl -H /dev/sda
Print Health Status of Disk Drive
smartctl -c /dev/sda
Print SMART capabilities that have been implemented
Monitoring Disk Health inorder to avoid data lost can be an every day task, but SMART has deamon smartd wich dose it for us:
lets enable and use it vi /etc/default/smartmontools :
1
# Defaults for smartmontools initscript (/etc/init.d/smartmontools)
2
# This is a POSIX shell fragment
3
​
4
# List of devices you want to explicitly enable S.M.A.R.T. for
5
# Not needed (and not recommended) if the device is monitored by smartd
6
#enable_smart="/dev/hda /dev/hdb"
7
​
8
# uncomment to start smartd on system startup
9
start_smartd=yes
10
​
11
# uncomment to pass additional options to smartd on startup
12
#smartd_opts="--interval=1800"
Copied!
and make sure smartd is running and enabled:
1
[email protected]:~# systemctl list-unit-files --type=service  | grep -i smart
2
smartd.service                             enabled 
3
smartmontools.service                      enabled 
4
[email protected]:~# systemctl status smartd
5
● smartd.service - Self Monitoring and Reporting Technology (SMART) Daemon
6
   Loaded: loaded (/lib/systemd/system/smartd.service; enabled; vendor preset: enabled)
7
   Active: active (running) since Mon 2018-01-01 13:35:51 +0330; 43min ago
8
     Docs: man:smartd(8)
9
           man:smartd.conf(5)
10
 Main PID: 4507 (smartd)
11
   CGroup: /system.slice/smartd.service
12
           └─4507 /usr/sbin/smartd -n
13
​
14
Jan 01 13:35:51 server1 smartd[4507]: Device: /dev/sda, type changed from 'scsi' to 'sat'
15
Jan 01 13:35:51 server1 smartd[4507]: Device: /dev/sda [SAT], opened
16
Jan 01 13:35:51 server1 smartd[4507]: Device: /dev/sda [SAT], Samsung SSD 850 EVO 500GB, S/N:S2R9NX0H503451W, WWN:5-002538-d40e515dc, FW:EMT02B6Q, 5
17
Jan 01 13:35:51 server1 smartd[4507]: Device: /dev/sda [SAT], found in smartd database: Samsung based SSDs
18
Jan 01 13:35:51 server1 smartd[4507]: Device: /dev/sda [SAT], can't monitor Current_Pending_Sector count - no Attribute 197
19
Jan 01 13:35:51 server1 smartd[4507]: Device: /dev/sda [SAT], can't monitor Offline_Uncorrectable count - no Attribute 198
20
Jan 01 13:35:51 server1 smartd[4507]: Device: /dev/sda [SAT], is SMART capable. Adding to "monitor" list.
21
Jan 01 13:35:51 server1 smartd[4507]: Monitoring 1 ATA and 0 SCSI devices
22
Jan 01 13:35:51 server1 smartd[4507]: Device: /dev/sda [SAT], state written to /var/lib/smartmontools/smartd.Samsung_SSD_850_EVO_500GB-S2R9NX0H50345
23
Jan 01 13:35:51 server1 systemd[1]: Started Self Monitoring and Reporting Technology (SMART) Daemon.
Copied!