01-14-2011
I have the same problem. I accidentally overwrote the drive with ZIL when upgrading from OpenSolaris 2009.11, to Solaris 11 Express (151a). Now I can't import the pool, even with the -m option (which documentation explicitly says it to do just that).
Sorry I don't have an answer for you (yet), but your effort - and more specifically your post - has at least save ME from potentially weeks of hair-pulling similar to what you have done.
I love the IDEA of ZFS, but in truth I have seen a significant "negative" benefit over the years of using it. I.E., loss of data. Solaris is so damn fickle and exceptionally fragile (as an operating system...and as I'm learning, ZFS is too). I mean, come on! The whole idea of an separate ZIL was to improve performance...and hopefully not at the expense of reliability! From now on, I will be disabling the ZIL altogether. (And the moment btrfs is production-ready on Linux, will migrate everything to that in a heartbeat.)
Who would have thought that losing the ZIL would render all of your data useless!?? Talk about unnecessary additional points of failure. (Even if mirrored...still an additional point of failure.) This risk, IMO, vastly outweighs the risk of asynchronous IO, unless you are running something that explicitly (i.e. documented) requires synchronous IO such as apps/databases that life and/or lots of $ depends on.
Anyway. I'm going to try to import via Linuz/ZFS-FUSE. If that fails I'll have to try to reconstruct the data via sporadic offline backups. I'll keep ya posted.
8 More Discussions You Might Find Interesting
1. Filesystems, Disks and Memory
Hi all,
I am looking at a MC cluster comprised of 2 nodes. There is a package named backup which does nothing more than mount a VG.
At one point we noticed errors in syslog that look like this:
cmclconfd: Could not access device file /dev/dsk/c20t0d0: No such file or directory
There are... (1 Reply)
Discussion started by: mariusp
1 Replies
2. Solaris
Hi,
I need to create 4 zones.
I have an array with a RAID5 volume. I have exported 4 LUNS from the array. Now my host see the 4 LUNS as raw devices.
I am partitioning each LUN using format then mounting each slice on the global zone. Then I am using each slice as a file system for my non... (3 Replies)
Discussion started by: general_lee
3 Replies
3. Red Hat
Hi All,
Could anyone please help to resolve the below problem.
I installed RHEL5.5 in my desktop.But when i try to activate the ethernet connection then it gives me the error.
I spent 2 days for the above and go through with several suggestion found by googling. But no luck.
... (0 Replies)
Discussion started by: Tanmoy
0 Replies
4. HP-UX
Hello,
I am following the HPUX 11.31 install/update guide and I am trying to install "Update-UX" from the installation media. I put the CD into the drive, and I am trying to mount the device. The instructions state:Find the DVD-ROM device file name:
ioscan -C disk -f -n -k | more
A typical... (5 Replies)
Discussion started by: bstring
5 Replies
5. Solaris
Weird duplicate device in one of my nested raidz1s - device is shown as both online and unavailable, with a hot spare showing up as degraded as well.
Any thoughts on how I should proceed?
root@storage# zpool status tank
pool: tank
state: DEGRADED
status: One or more devices are... (6 Replies)
Discussion started by: DoohanMcGirk
6 Replies
6. Solaris
A backup/clone script of ours was recently ran. It normally only clones the rpool and renames in rpoolA. Something must've changed as it found another one of our pools that it shouldn't have. It exported that pool unbeknownst to us. Later on when a coworker realized the other pool was missing he... (2 Replies)
Discussion started by: beantownmp
2 Replies
7. HP-UX
Hi,
Unable to make tape backup, please help.
/opt/ignite/bin/make_tape_recovery -a /dev/rmt/?mn -I -v -m tar -x inc_entire=vg00
* Creating local directories for configuration files and archive.
======= 04/25/16 16:28:08 IST Started /opt/ignite/bin/make_tape_recovery.
(Mon... (4 Replies)
Discussion started by: anuragr
4 Replies
8. Solaris
On an OmniOS server, I removed a single-disk pool I was using for testing.
Now, when I run zpool import it will show it as FAULTED, since that single disk not available anymore.
# zpool import
pool: fido
id: 7452075738474086658
state: FAULTED
status: The pool was last... (11 Replies)
Discussion started by: priyadarshan
11 Replies
LEARN ABOUT CENTOS
amzfs-sendrecv
AMZFS-SENDRECV(8) System Administration Commands AMZFS-SENDRECV(8)
NAME
amzfs-sendrecv - Amanda script to create zfs sendrecv
DESCRIPTION
amzfs-sendrecv is an Amanda application implementing the Application API. It should not be run by users directly. It create a zfs snapshot
of the filesystem and backup the snapshot with 'zfs send'. Snapshot are kept after the backup is done, this increase the disk space use on
the client but it is neccesary to be able do to incremental backup. If you want only full backup, you can disable this feature by setting
the KEEP-SNAPSHOT property to 'NO'. Only the restoration of the complete backup is allowed, it is impossible to restore a single file.
The application is run as the amanda user, it must have many zfs priviledge:
zfs allow -ldu AMANDA_USER mount,create,rename,snapshot,destroy,send,receive FILESYSTEM
Some system doesn't have "zfs allow", but you can give the Amanda backup user the rights to manipulate ZFS filesystems by using the
following command:
usermod -P "ZFS File System Management,ZFS Storage Management" AMANDA_USER
This will require that your run zfs under pfexec, set the PFEXEC property to YES.
The format of the diskdevice in the disklist (DLE) must be one of:
Desciption Example
---------- -------
Mountpoint /data
ZFS pool name datapool
ZFS filesystem datapool/database
ZFS logical volume datapool/dbvol
The filesystem doesn't need to be mounted.
PROPERTIES
This section lists the properties that control amzfs-sendrecv's functionality. See amanda-applications(7) for information on the
Application API, application configuration.
DF-PATH
Path to the 'df' binary, search in $PATH by default.
KEEP-SNAPSHOT
If "YES" (the default), snapshot are kept after the backup, if set to "NO" then snapshot are no kept and incremental backup will fail.
ZFS-PATH
Path to the 'zfs' binary, search in $PATH by default.
PFEXEC-PATH
Path to the 'pfexec' binary, search in $PATH by default.
PFEXEC
If "NO" (the default), pfexec is not used, if set to "YES" then pfexec is used.
EXAMPLE
In this example, a dumptype is defined to use amzfs-sendrecv application to backup a zfs filesystem.
define application-tool amzfs_sendrecv {
comment "amzfs-sendrecv"
plugin "amzfs-sendrecv"
#property "DF-PATH" "/usr/sbin/df"
#property "KEEP-SNAPSHOT" "YES"
#property "ZFS-PATH" "/usr/sbin/zfs"
#property "PFEXEC-PATH" "/usr/sbin/pfexec"
#property "PFEXEC" "NO"
}
define dumptype user-zfs-sendrecv {
program "APPLICATION"
application "amzfs_sendrecv"
}
SEE ALSO
amanda(8), amanda.conf(5), amanda-client.conf(5), amanda-applications(7)
The Amanda Wiki: : http://wiki.zmanda.com/
AUTHOR
Jean-Louis Martineau <martineau@zmanda.com>
Zmanda, Inc. (http://www.zmanda.com)
Amanda 3.3.3 01/10/2013 AMZFS-SENDRECV(8)