ZFS -- the first sip of the world best file system --
Sunday Nov 12, 2006
I have a chance to test the Solaris 10 06/06 ZFS features. I found that it is very simple to config and admin. In just one CLI command, it is ready to use. This is the simplest file system I have ever used. ZFS is the best.
Listing below are some features for ZFS in Solaris 10 Update 2 version :
1. World first 128-bit filesystem
2. 99.99999999999999999 integrity
3. Self healing file system -- no fsck (filesystem check needed)
4. Architect for Integrity and Speed
5. Rock Solid availability with RAID-Z (Mirroring also supported)
6. No Volume management software needed (uses Storage Pool concept instead)
7. Dynamic Stripping (to maximize disk spindles)
8. Built-in data services such as Snapshot and Cloning
9. Quota and Capacity Reservation supported
10. Data Compression supported
11. Data Encryption supported (in future release)
12. Very Strong Integrity model --- Everything is Transactional, Copy-on-Write (COW) and End-to-End Checksummed. Help prevent from Bit rot, Phantom writes, Misdirected reads and writes, H/W and DMA parity errors, S/W and Driver bugs, Accidental overwrite
13. Resilver and Resynchronization
14. Very simple to admin (via CLI and GUI)
for CLI, very few commands needed
15. ETC ...
Below is my screen captured from my notebook when testing Solaris 10 x86 (06/06) for ZFS feature:
(1) Show physical disk or file disk type for ZFS
#ls -la /zfsdisk/*
-rw------T 1 root root 104857600 May 17 14:39 /zfsdisk/disk1
-rw------T 1 root root 104857600 May 17 14:39 /zfsdisk/disk2
-rw------T 1 root root 104857600 May 17 14:42 /zfsdisk/disk3
-rw------T 1 root root 104857600 May 17 14:23 /zfsdisk/disk4
-rw------T 1 root root 157286400 May 17 14:37 /zfsdisk/disk5
-rw------T 1 root root 157286400 May 17 14:42 /zfsdisk/disk6
NOTIECE : disk1-4 have 100MB in size, disk5-6 have 150MB.
(2) Initial creation of pool (mypool) -- zpool create
# zpool create mypool mirror /zfsdisk/disk1 /zfsdisk/disk2
# zpool list
NAME SIZE USED AVAIL CAP HEALTH ALTROOT
mypool 95.5M 52.5K 95.4M 0% ONLINE -
# zpool status
pool: mypool
state: ONLINE
scrub: none requested
config:
NAME STATE READ WRITE CKSUM
mypool ONLINE 0 0 0
mirror ONLINE 0 0 0
/zfsdisk/disk1 ONLINE 0 0 0
/zfsdisk/disk2 ONLINE 0 0 0
errors: No known data errors
(3) Initial creation of file systems (myfs) -- zfs create
# zfs create mypool/myfs
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 104K 63.4M 24.5K /mypool
mypool/myfs 24.5K 63.4M 24.5K /mypool/myfs
(4) Adding extra disks -- zpool add
# zpool add mypool mirror /zfsdisk/disk3 /zfsdisk/disk4
# zpool status
pool: mypool
state: ONLINE
scrub: none requested
config:
NAME STATE READ WRITE CKSUM
mypool ONLINE 0 0 0
mirror ONLINE 0 0 0
/zfsdisk/disk1 ONLINE 0 0 0
/zfsdisk/disk2 ONLINE 0 0 0
mirror ONLINE 0 0 0
/zfsdisk/disk3 ONLINE 0 0 0
/zfsdisk/disk4 ONLINE 0 0 0
errors: No known data errors
#zpool list
NAME SIZE USED AVAIL CAP HEALTH ALTROOT
mypool 191M 214K 191M 0% ONLINE -
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 106K 159M 25.5K /mypool
mypool/myfs 24.5K 159M 24.5K /mypool/myfs
*** Avail size from the 'zpool list' and 'zfs list' commands may vary slightly
*** as the 'zfs list' command accounts for a small amount of space reserved for
*** the file system level operations that is NOT visible from the 'zpool list'
*** command
(5) Volumes
# zfs create -V 50m mypool/myvol
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 50.1M 109M 25.5K /mypool
mypool/myfs 24.5K 109M 24.5K /mypool/myfs
mypool/myvol 22.5K 159M 22.5K -
*** no mount point listed for myvol, Volumes are NOT mountable
*** (need to make it UFS to be able to mount)
*** ZFS has guaranteed that there will be 50MB available -
*** the other file systems available space has been reduced by
*** 50MB and the pool has 50MB used
# newfs /dev/zvol/rdsk/mypool/myvol
newfs: construct a new file system /dev/zvol/rdsk/mypool/myvol: (y/n)? y
Warning: 2048 sector(s) in last cylinder unallocated
/dev/zvol/rdsk/mypool/myvol: 102400 sectors in 17 cylinders of 48 tracks, 128 sectors
50.0MB in 2 cyl groups (14 c/g, 42.00MB/g, 20160 i/g)
super-block backups (for fsck -F ufs -o b=#) at:
32, 86176,
# mkdir /data
# mount /dev/zvol/dsk/mypool/myvol /data
# df -k
Filesystem kbytes used avail capacity Mounted on
/dev/dsk/c0d0s0 14435859 3730629 10560872 27% /
/devices 0 0 0 0% /devices
ctfs 0 0 0 0% /system/contract
proc 0 0 0 0% /proc
mnttab 0 0 0 0% /etc/mnttab
swap 2734616 700 2733916 1% /etc/svc/volatile
objfs 0 0 0 0% /system/object
/usr/lib/libc/libc_hwcap1.so.1
14435859 3730629 10560872 27% /lib/libc.so.1
fd 0 0 0 0% /dev/fd
swap 2733972 56 2733916 1% /tmp
swap 2733940 24 2733916 1% /var/run
/vol/dev/dsk/c1t0d0/sol_10_606_x86
362602 362602 0 100% /cdrom/sol_10_606_x86
mypool 162816 25 111472 1% /mypool
mypool/myfs 162816 24 111472 1% /mypool/myfs
/dev/zvol/dsk/mypool/myvol
46111 1041 40459 3% /data
*** myvol has used aprox 5MB.
(6) Destroying myvol
# zfs destroy mypool/myvol
(7) Additional file system
# zpool create mypool/myfs2
(8) Reservations
(to guaratee that a file system has a certain level of capacity available to it)
*** do this AFTER you have created multiple file systems but
*** BEFORE demo the quota section
# zfs set reservation=155M mypool/myfs
# zfs get reservation mypool/myfs
NAME PROPERTY VALUE SOURCE
mypool/myfs reservation 155M local
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 155M 3.86M 27.5K /mypool
mypool/myfs 24.5K 159M 24.5K /mypool/myfs
mypool/myfs2 24.5K 3.86M 24.5K /mypool/myfs2
***
*** copy data over with in the limits of the space available
***
# df -k
Filesystem kbytes used avail capacity Mounted on
/dev/dsk/c0d0s0 14435859 3729421 10562080 27% /
. . . . .
. . . . .
mypool 162816 27 3950 1% /mypool
mypool/myfs 162816 24 162645 1% /mypool/myfs
mypool/myfs2 162816 24 3950 1% /mypool/myfs2
# ls -la /kernel/genunix
-rwxr-xr-x 1 root sys 2235280 Apr 30 05:10 /kernel/genunix
# cp /kernel/genunix /mypool/myfs2/genunix1
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 157M 1.60M 27.5K /mypool
mypool/myfs 24.5K 157M 24.5K /mypool/myfs
mypool/myfs2 2.28M 1.60M 2.28M /mypool/myfs2
# df -k
. . . . .
mypool 162816 27 1640 2% /mypool
mypool/myfs 162816 24 160336 1% /mypool/myfs
mypool/myfs2 162816 2331 1640 59% /mypool/myfs2
# cp /kernel/genunix /mypool/myfs2/genunix2
cp: /mypool/myfs2/genunix2: No space left on device
*** ZFS actually rolls back the data that was copied cover so you will NOT
*** hit the full file system situation.
(9) Reservation -- unset
# zfs set reservation=none mypool/myfs
(10) Quotas -- maximum limit on the size a file system can be used
# zfs set quota=2m mypool/myfs2
# zfs get quota mypool/myfs2
NAME PROPERTY VALUE SOURCE
mypool/myfs2 quota 2M local
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 174K 159M 27.5K /mypool
mypool/myfs 24.5K 159M 24.5K /mypool/myfs
mypool/myfs2 24.5K 1.98M 24.5K /mypool/myfs2
# cp /kernel/genunix /mypool/myfs2/genunix1
cp: /kernel/genunix: Disc quota exceeded
*** Like the 'Reservation', ZFS undoes the effects of the copy as it did not
*** complete
(11) Auto NFS Sharing
# share
< no output from the share command>
# zfs set sharenfs=on mypool/myfs
# zfs get sharenfs mypool/myfs
NAME PROPERTY VALUE SOURCE
mypool/myfs sharenfs on local
# share
- /mypool/myfs rw ""
----Unshare----
# zfs set sharenfs=off mypool/myfs
# zfs get sharenfs mypool/myfs
NAME PROPERTY VALUE SOURCE
mypool/myfs sharenfs off local
#
# share
< no output from the share command>
(12) Data Recovery -- Shapshots (creating Write-protected image)
----copy data to myfs----
# cp /usr/dict/words /mypool/myfs
# cp /etc/passwd /mypool/myfs
# ls -la /mypool/myfs
total 526
drwxr-xr-x 2 root sys 4 May 17 16:07 .
drwxr-xr-x 4 root sys 4 May 17 15:38 ..
-rw-r--r-- 1 root root 671 May 17 16:07 passwd
-r--r--r-- 1 root root 206663 May 17 16:07 words
----take a snapshot of myfs----
# zfs snapshot mypool/myfs@first
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 443K 159M 27.5K /mypool
mypool/myfs 284K 159M 284K /mypool/myfs
mypool/myfs@first 0 - 284K -
mypool/myfs2 24.5K 1.98M 24.5K /mypool/myfs2
----take a snapshot of myvol----
# zfs snapshot mypool/myvol@backup
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 50.1M 13.4M 25.5K /mypool
mypool/myfs 24.5K 13.4M 24.5K /mypool/myfs
mypool/myvol 22.5K 63.4M 22.5K -
mypool/myvol@backup 0 - 22.5K -
----verify snapshot----
ls -la /mypool/myfs/.zfs/snapshot/first
total 523
drwxr-xr-x 2 root sys 4 May 17 16:07 .
dr-xr-xr-x 3 root root 3 May 17 14:55 ..
-rw-r--r-- 1 root root 671 May 17 16:07 passwd
-r--r--r-- 1 root root 206663 May 17 16:07 words
*** if we created mypool/myvol (zfs create -V 50m mypool/myvol), then we
*** made it UFS file system (newfs /dev/zvol/rdsk/mypool/myvol), then we
*** mounted this /dev/zvol/dsk/mypool/myvol to /data
*** we can do the snapshot for this volume as well, by using this command
*** zfs snapshot mypool/myvol@first, the /dev/zvol/dsk/mypool@first can be
*** mounted with RO (read-only)
*** if we snapshot ZFS using 'zfs snapshot mypool/myfs@backup', for example;
*** the snapshot will be located in '/mypool/myfs/.zfs/snapshot/backup'
# cd /mypool/myfs
# mkfile -n 5m junk
# ls -la
total 527
drwxr-xr-x 2 root sys 5 May 17 16:14 .
drwxr-xr-x 4 root sys 4 May 17 15:38 ..
-rw------T 1 root root 5242880 May 17 16:14 junk
-rw-r--r-- 1 root root 671 May 17 16:07 passwd
-r--r--r-- 1 root root 206663 May 17 16:07 words
# df -k
Filesystem kbytes used avail capacity Mounted on
. . . . .
mypool 162816 27 162210 1% /mypool
mypool/myfs 162816 414 162210 1% /mypool/myfs
mypool/myfs2 2048 24 2023 2% /mypool/myfs2
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 606K 158M 27.5K /mypool
mypool/myfs 438K 158M 414K /mypool/myfs
mypool/myfs@first 23.5K - 284K -
mypool/myfs2 24.5K 1.98M 24.5K /mypool/myfs2
# rm junk
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 476K 159M 27.5K /mypool
mypool/myfs 308K 159M 284K /mypool/myfs
mypool/myfs@first 23.5K - 284K -
mypool/myfs2 24.5K 1.98M 24.5K /mypool/myfs2
*** now mypool/myfs && mypool/myfs@first has the same REFER side
*** now some lines from /mypool/myfs/words, then save
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 608K 158M 27.5K /mypool
mypool/myfs 440K 158M 156K /mypool/myfs
mypool/myfs@first 284K - 284K -
mypool/myfs2 24.5K 1.98M 24.5K /mypool/myfs2
*** now REFER side of mypool/myfs && mypool/myfs@first are NOT the same
#diff /mypool/myfs/words /mypool/myfs/.zfs/snapshot/first/words
<diff does some output>
---Data comparision---
# digest -a md5 /mypool/myfs/words
34a2d6e3c4851ea9a56fc5ace4ef7380 (CORRUPTED!!!)
#
# digest -a md5 /mypool/myfs/.zfs/snapshot/first/words
5dc66244a7bef7d3018538e144e4bbdc
---Roll back----
# zfs rollback mypool/myfs@first
# digest -a md5 /mypool/myfs/words
5dc66244a7bef7d3018538e144e4bbdc
# digest -a md5 /mypool/myfs/.zfs/snapshot/first/words
5dc66244a7bef7d3018538e144e4bbdc
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 444K 159M 27.5K /mypool
mypool/myfs 284K 159M 284K /mypool/myfs
mypool/myfs@first 0 - 284K -
mypool/myfs2 24.5K 1.98M 24.5K /mypool/myfs2
# diff /mypool/myfs/words /mypool/myfs/.zfs/snapshot/first/words
< no output here >
----Destroying the snapshot----
## zfs destroy mypool/myfs
cannot destroy 'mypool/myfs': filesystem has children
use '-r' to destroy the following datasets:
mypool/myfs@first
# zfs destroy mypool/myfs@first
(13) Import/Export
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 444K 159M 27.5K /mypool
mypool/myfs 284K 159M 284K /mypool/myfs
mypool/myfs2 24.5K 1.98M 24.5K /mypool/myfs2
# zpool list
NAME SIZE USED AVAIL CAP HEALTH ALTROOT
mypool 191M 458K 191M 0% ONLINE -
# zpool export -f mypool
# zpool list
no pools available
# zfs list
no datasets available
(14) Backup -- HAVEN'T TESTED YET
(15) Device Replacement
In case, we pretend that /zfsdisk/disk4 was corrupted, we will replace it with
/zfsdisk/disk6, by using 'zpool replace' command
# zpool status
pool: mypool
state: ONLINE
scrub: none requested
config:
NAME STATE READ WRITE CKSUM
mypool ONLINE 0 0 0
raidz ONLINE 0 0 0
/zfsdisk/disk1 ONLINE 0 0 0
/zfsdisk/disk2 ONLINE 0 0 0
/zfsdisk/disk3 ONLINE 0 0 0
/zfsdisk/disk4 ONLINE 0 0 0
/zfsdisk/disk5 ONLINE 0 0 0
errors: No known data errors
# zpool replace mypool /zfsdisk/disk4 /zfsdisk/disk6
# zpool status
pool: mypool
state: DEGRADED
status: One or more devices is currently being resilvered. The pool will
continue to function, possibly in a degraded state.
action: Wait for the resilver to complete.
scrub: resilver in progress, 0.39% done, 0h12m to go
config:
NAME STATE READ WRITE CKSUM
mypool DEGRADED 0 0 0
raidz DEGRADED 0 0 0
/zfsdisk/disk1 ONLINE 0 0 0
/zfsdisk/disk2 ONLINE 0 0 0
/zfsdisk/disk3 ONLINE 0 0 0
replacing DEGRADED 0 0 0
/zfsdisk/disk4 UNAVAIL 0 0 0 corrupted data
/zfsdisk/disk6 ONLINE 0 0 0
/zfsdisk/disk5 ONLINE 0 0 0
errors: No known data errors
***
*** Only the actual data in use in file systems, snapshots, zvols etc will be
*** resilvered, not the entire file systems, snapshots, zvols
***
***
*** NOTE
***
*** the /zfsdisk/disk4 size is 100MB, the /zfsdisk/disk6 size is 150MB
*** the drive disk6 can replace disk4, but disk4 can not replace disk6
*** For example,
** # zpool replace mypool /zfsdisk/disk6 /zfsdisk/disk4
**cannot replace /zfsdisk/disk6 with /zfsdisk/disk4: /zfsdisk/disk4 is too small
(16) Scrubbing the pool -- to check the validity of the specified ZFS checksums
# zpool status
pool: mypool
state: ONLINE
scrub: resilver completed with 0 errors on Thu May 18 14:28:38 2006
config:
NAME STATE READ WRITE CKSUM
mypool ONLINE 0 0 0
raidz ONLINE 0 0 0
/zfsdisk/disk1 ONLINE 0 0 0
/zfsdisk/disk2 ONLINE 0 0 0
/zfsdisk/disk3 ONLINE 0 0 0
/zfsdisk/disk6 ONLINE 0 0 0
/zfsdisk/disk5 ONLINE 0 0 0
errors: No known data errors
# zpool scrub mypool
# zpool status
pool: mypool
state: ONLINE
scrub: scrub in progress, 18.13% done, 0h0m to go
config:
NAME STATE READ WRITE CKSUM
mypool ONLINE 0 0 0
raidz ONLINE 0 0 0
/zfsdisk/disk1 ONLINE 0 0 0
/zfsdisk/disk2 ONLINE 0 0 0
/zfsdisk/disk3 ONLINE 0 0 0
/zfsdisk/disk6 ONLINE 0 0 0
/zfsdisk/disk5 ONLINE 0 0 0
errors: No known data errors
***
*** Scrubing is done 18.13%
***
# zpool status
pool: mypool
state: ONLINE
scrub: scrub completed with 0 errors on Thu May 18 14:39:15 2006
config:
NAME STATE READ WRITE CKSUM
mypool ONLINE 0 0 0
raidz ONLINE 0 0 0
/zfsdisk/disk1 ONLINE 0 0 0
/zfsdisk/disk2 ONLINE 0 0 0
/zfsdisk/disk3 ONLINE 0 0 0
/zfsdisk/disk6 ONLINE 0 0 0
/zfsdisk/disk5 ONLINE 0 0 0
errors: No known data errors
***
*** Scrubing is done with 0 errors
***
(17) ZFS Compression
*** This will compress only ZFS file system created by
*** created by 'zfs create mypool/myfs', for example
***
*** this will NOT work with UFS volume created
*** by 'zfs create -V 50M mypool/myvol',then 'newfs /dev/zvol/rdsk/mypool/myvol'
***
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 131M 313M 51K /mypool
mypool/myfs 49K 313M 49K /mypool/myfs
mypool/myvol 131M 313M 113M -
mypool/myvol@backup 18.3M - 126M -
# cd /mypool/myfs
# zfs get compression mypool/myfs
NAME PROPERTY VALUE SOURCE
mypool/myfs compression off local
# zfs get compressratio mypool/myfs
NAME PROPERTY VALUE SOURCE
mypool/myfs compressratio 1.00x -
# cp /kernel/genunix /mypool/myfs
# du -k /mypool/myfs/genunix
2884 /mypool/myfs/genunix
---- Set compression to ON ----
# zfs set compression=on mypool/myfs
# zfs get compression mypool/myfs
NAME PROPERTY VALUE SOURCE
mypool/myfs compression on local
# zfs get compressratio mypool/myfs
NAME PROPERTY VALUE SOURCE
mypool/myfs compressratio 1.00x -
***
*** Enabling compression will ONLY effect data written after this point,
*** it is not applied retrospectively
***
# cp /kernel/genunix /mypool/myfs/genunix_compressed
# zfs get compressratio mypool/myfs
NAME PROPERTY VALUE SOURCE
mypool/myfs compressratio 1.21x -
# du -k /mypool/myfs/gen*
2884 /mypool/myfs/genunix
1865 /mypool/myfs/genunix_compressed
*** if later we set the ZFS compression to OFF, the already compressed files
*** are still compressed, but the later written files to this /mypool/myfs,
*** will not be further compressed...
(18) Data Recovery -- Clones (creating of writable image)
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 113M 331M 51K /mypool
mypool/myfs 49K 331M 49K /mypool/myfs
mypool/myvol 113M 331M 113M -
*** mypool/myfs used only 49KB
# for i in 1 2 3 4 5 6 7 8 9
> do
> cp /kernel/genunix /mypool/myfs/genunix-$i
> done
# ls -la /mypool/myfs/gen*
-rwxr-xr-x 1 root root 2235280 May 18 15:14 /mypool/myfs/genunix-1
-rwxr-xr-x 1 root root 2235280 May 18 15:14 /mypool/myfs/genunix-2
-rwxr-xr-x 1 root root 2235280 May 18 15:14 /mypool/myfs/genunix-3
-rwxr-xr-x 1 root root 2235280 May 18 15:14 /mypool/myfs/genunix-4
-rwxr-xr-x 1 root root 2235280 May 18 15:14 /mypool/myfs/genunix-5
-rwxr-xr-x 1 root root 2235280 May 18 15:14 /mypool/myfs/genunix-6
-rwxr-xr-x 1 root root 2235280 May 18 15:14 /mypool/myfs/genunix-7
-rwxr-xr-x 1 root root 2235280 May 18 15:14 /mypool/myfs/genunix-8
-rwxr-xr-x 1 root root 2235280 May 18 15:14 /mypool/myfs/genunix-9
# digest -a md5 /mypool/myfs/gen*
(/mypool/myfs/genunix-1) = 6a594ec25150b6b84b4313c9f111bad2
(/mypool/myfs/genunix-2) = 6a594ec25150b6b84b4313c9f111bad2
(/mypool/myfs/genunix-3) = 6a594ec25150b6b84b4313c9f111bad2
(/mypool/myfs/genunix-4) = 6a594ec25150b6b84b4313c9f111bad2
(/mypool/myfs/genunix-5) = 6a594ec25150b6b84b4313c9f111bad2
(/mypool/myfs/genunix-6) = 6a594ec25150b6b84b4313c9f111bad2
(/mypool/myfs/genunix-7) = 6a594ec25150b6b84b4313c9f111bad2
(/mypool/myfs/genunix-8) = 6a594ec25150b6b84b4313c9f111bad2
(/mypool/myfs/genunix-9) = 6a594ec25150b6b84b4313c9f111bad2
#
# zfs snapshot mypool/myfs@forclone
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 138M 306M 51K /mypool
mypool/myfs 25.4M 306M 25.4M /mypool/myfs
mypool/myfs@forclone 0 - 25.4M -
mypool/myvol 113M 306M 113M -
*** mypool/myfs now used 25.4MB, mypool@myfs@forclone also REFERS to 25.4MB
*** but occupied 0MB.
---- Creating 2 clone images named mypool/clone1 & mypool/clone2 ----
# zfs clone mypool/myfs@forclone mypool/clone1
# zfs clone mypool/myfs@forclone mypool/clone2
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 138M 306M 53K /mypool
mypool/clone1 0 306M 25.4M /mypool/clone1
mypool/clone2 0 306M 25.4M /mypool/clone2
mypool/myfs 25.4M 306M 25.4M /mypool/myfs
mypool/myfs@forclone 0 - 25.4M -
mypool/myvol 113M 306M 113M -
# cd /mypool
# du -k
25963 ./clone2
25963 ./myfs
25963 ./clone1
77891 .
---- Make change in the original and in the first clone (clone1) ---
# mkfile 1m /mypool/myfs/clonetest
# mkfile 2m /mypool/clone1/clonetest
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 142M 302M 55K /mypool
mypool/clone1 2.55M 302M 27.9M /mypool/clone1
mypool/clone2 0 302M 25.4M /mypool/clone2
mypool/myfs 26.7M 302M 26.7M /mypool/myfs
mypool/myfs@forclone 49K - 25.4M -
mypool/myvol 113M 302M 113M -
*** mypool/myfs used up more space (from 25.4MB to 26.7MB)
*** mypool/clone1 used up more space (from 0MB to 2.55MB)
---- Recursively removing the snapshot and the clones that use it ---
---- in this case, will recursively delete mypool/myfs@forclone,
---- mypool/clone1 and mypool/clone2, since clone1 & clone2 is using
---- mypool/myfs@forclone
----
---- @ will be used with snapshot
----
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 142M 302M 55K /mypool
mypool/clone1 2.55M 302M 27.9M /mypool/clone1
mypool/clone2 0 302M 25.4M /mypool/clone2
mypool/myfs 26.7M 302M 26.7M /mypool/myfs
mypool/myfs@forclone 49K - 25.4M -
mypool/myvol 113M 302M 113M -
# zfs destroy -R mypool/myfs@forclone
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 140M 304M 51K /mypool
mypool/myfs 26.7M 304M 26.7M /mypool/myfs
mypool/myvol 113M 304M 113M -
(19) Property inheritance
to inherit properties to multiple file systems at the same time
# zfs create mypool/homedirs
# zfs create mypool/homedirs/user1
# zfs create mypool/homedirs/user2
# zfs create mypool/homedirs/user3
# zfs list
NAME USED AVAIL REFER MOUNTPOINT
mypool 140M 304M 53K /mypool
mypool/homedirs 200K 304M 53K /mypool/homedirs
mypool/homedirs/user1 49K 304M 49K /mypool/homedirs/user1
mypool/homedirs/user2 49K 304M 49K /mypool/homedirs/user2
mypool/homedirs/user3 49K 304M 49K /mypool/homedirs/user3
mypool/myfs 26.7M 304M 26.7M /mypool/myfs
mypool/myvol 113M 304M 113M -
# zfs get compression mypool/homedirs
NAME PROPERTY VALUE SOURCE
mypool/homedirs compression off default
# zfs get compression mypool/homedirs/user1
NAME PROPERTY VALUE SOURCE
mypool/homedirs/user1 compression off default
#
# zfs set compression=on mypool/homedirs/user1
# zfs get compression mypool/homedirs
NAME PROPERTY VALUE SOURCE
mypool/homedirs compression off default
# zfs get compression mypool/homedirs/user1
NAME PROPERTY VALUE SOURCE
mypool/homedirs/user1 compression on local
# zfs set compression=on mypool/homedirs
# zfs get compression mypool/homedirs/user1 mypool/homedirs/user2 mypool/homedirs/user3NAME PROPERTY VALUE SOURCE
mypool/homedirs/user1 compression on local
mypool/homedirs/user2 compression on inherited from mypool/homedirs
mypool/homedirs/user3 compression on inherited from mypool/homedirs
*** mypool/homedirs/user2 & mypool/homedirs/user3 inherited compression=on
*** from mypool/homedirs, but mypool/homedirs/user1 was set compression=on
*** independently (stated local)
---- set mypool/homedirs/user1 to inherit compression property from its
---- upper mypool/homedirs file system
# zfs inherit compression mypool/homedirs/user1
# zfs get compression mypool/homedirs/user1 mypool/homedirs/user2 mypool/homedirs/user3
NAME PROPERTY VALUE SOURCE
mypool/homedirs/user1 compression on inherited from mypool/homedirs
mypool/homedirs/user2 compression on inherited from mypool/homedirs
mypool/homedirs/user3 compression on inherited from mypool/homedirs
---- check whether which properties are inheritable ----
# zfs get (see INHERIT)
. . . . . .
. . . . . .
PROPERTY EDIT INHERIT VALUES
type NO NO filesystem | volume | snapshot
creation NO NO <date>
used NO NO <size>
available NO NO <size>
referenced NO NO <size>
compressratio NO NO <1.00x or higher if compressed>
mounted NO NO yes | no | -
origin NO NO <snapshot>
quota YES NO <size> | none
reservation YES NO <size> | none
volsize YES NO <size>
volblocksize NO NO 512 to 128k, power of 2
recordsize YES YES 512 to 128k, power of 2
mountpoint YES YES <path> | legacy | none
sharenfs YES YES on | off | share(1M) options
checksum YES YES on | off | fletcher2 | fletcher4 | sha256
compression YES YES on | off | lzjb
atime YES YES on | off
devices YES YES on | off
exec YES YES on | off
setuid YES YES on | off
readonly YES YES on | off
zoned YES YES on | off
snapdir YES YES hidden | visible
aclmode YES YES discard | groupmask | passthrough
aclinherit&a
|
|
|
|
Posted by Armad Mamu on March 11, 2007 at 03:58 PM ICT #