Discussion:
zpool swap errors
(too old to reply)
Nikhil
2010-08-24 07:39:13 UTC
Permalink
Hi,

I have a netbooted node on which I am using a zfspool built with
mirror configuration on top of the two disks for arbitrary storage. On
top of the zfspool, the zfs swap filesystem is also used. Once in a
while I keep running into the problem where the zpool status -v
reports problems with swap filesystem and following is the depiction
of the error.

=================================================================
# zpool status -v
pool: zpool0
state: ONLINE
status: One or more devices has experienced an error resulting in data
corruption. Applications may be affected.
action: Restore the file in question if possible. Otherwise restore
the
entire pool from backup.
see: http://www.sun.com/msg/ZFS-8000-8A
scrub: none requested
config:

NAME STATE READ WRITE CKSUM
zpool0 ONLINE 0 0 4
c0t0d0 ONLINE 0 0 4

errors: Permanent errors have been detected in the following files:

zpool0/swap:<0x1>
=================================================================

Since the machine is a netbooted and I configure the zfspool everytime
the machine reboots , this problem can get resolved since the
filesystem is going to be also newly created.

But Apart from the reboot, I am not finding any solution that can be
used to fix this problem while keeping the host online without
disturbing any of the applications.

Is there a way that will follow what the action item from the status
summary tells, as in finding the file in question? and the file in
question is in a no way of clue as the file is shown in hex.

What is the solution to this problem? Why does this keep occurring in
its first place? How can it be avoided? How can it be fixed?

Appreciate your knowledge transferr on this.

Thanks,
Nikhil
Ian Collins
2010-08-24 07:54:07 UTC
Permalink
Post by Nikhil
Hi,
I have a netbooted node on which I am using a zfspool built with
mirror configuration on top of the two disks for arbitrary storage. On
top of the zfspool, the zfs swap filesystem is also used. Once in a
while I keep running into the problem where the zpool status -v
reports problems with swap filesystem and following is the depiction
of the error.
=================================================================
# zpool status -v
pool: zpool0
state: ONLINE
status: One or more devices has experienced an error resulting in data
corruption. Applications may be affected.
action: Restore the file in question if possible. Otherwise restore
the
entire pool from backup.
see: http://www.sun.com/msg/ZFS-8000-8A
scrub: none requested
NAME STATE READ WRITE CKSUM
zpool0 ONLINE 0 0 4
c0t0d0 ONLINE 0 0 4
You have a pool without any redundancy, never a good idea.
--
Ian Collins
Nikhil
2010-08-24 08:18:27 UTC
Permalink
Post by Ian Collins
Post by Nikhil
Hi,
I have a netbooted node on which I am using a zfspool built with
mirror configuration on top of the two disks for arbitrary storage. On
top of the zfspool, the zfs swap filesystem is also used. Once in a
while I keep running into the problem where the zpool status -v
reports problems with swap filesystem and following is the depiction
of the error.
=================================================================
# zpool status -v
   pool: zpool0
  state: ONLINE
status: One or more devices has experienced an error resulting in data
         corruption.  Applications may be affected.
action: Restore the file in question if possible.  Otherwise restore
the
         entire pool from backup.
    see:http://www.sun.com/msg/ZFS-8000-8A
  scrub: none requested
         NAME        STATE     READ WRITE CKSUM
         zpool0         ONLINE       0     0     4
           c0t0d0    ONLINE       0     0     4
You have a pool without any redundancy, never a good idea.
--
Ian Collins- Hide quoted text -
- Show quoted text -
Hey Ian,

I did not get that...pool without any redundancy? pool has two disks
with mirror configuration and this is basically for arbitrary storage
only so I do not really be using the redundant pool. (Is not the pool
by it self redundant in zfs nature?)

the swap is the only filesystem that runs into the problem.. so you
meant to say that zpool redundancy could be related? but how?
Ian Collins
2010-08-24 08:27:58 UTC
Permalink
Post by Nikhil
Post by Ian Collins
Post by Nikhil
Hi,
I have a netbooted node on which I am using a zfspool built with
mirror configuration on top of the two disks for arbitrary storage. On
top of the zfspool, the zfs swap filesystem is also used. Once in a
while I keep running into the problem where the zpool status -v
reports problems with swap filesystem and following is the depiction
of the error.
=================================================================
# zpool status -v
pool: zpool0
state: ONLINE
status: One or more devices has experienced an error resulting in data
corruption. Applications may be affected.
action: Restore the file in question if possible. Otherwise restore
the
entire pool from backup.
see:http://www.sun.com/msg/ZFS-8000-8A
scrub: none requested
NAME STATE READ WRITE CKSUM
zpool0 ONLINE 0 0 4
c0t0d0 ONLINE 0 0 4
You have a pool without any redundancy, never a good idea.
Hey Ian,
I did not get that...pool without any redundancy? pool has two disks
with mirror configuration and this is basically for arbitrary storage
only so I do not really be using the redundant pool. (Is not the pool
by it self redundant in zfs nature?)
The status list above only shows c0t0d0. Unlike other filesystems, ZFS
does not silently ignore errors. If you don't have redundancy, you can
easily end up with a corrupt pool.

Setting the "copies" property to 2 may help, but if you keep getting
errors you most likely have either bad memory, or a bad drive.
Post by Nikhil
the swap is the only filesystem that runs into the problem.. so you
meant to say that zpool redundancy could be related? but how?
What other filesystems are on the pool?
--
Ian Collins
Nikhil
2010-08-24 09:11:12 UTC
Permalink
Post by Nikhil
Post by Ian Collins
Post by Nikhil
Hi,
I have a netbooted node on which I am using a zfspool built with
mirror configuration on top of the two disks for arbitrary storage. On
top of the zfspool, the zfs swap filesystem is also used. Once in a
while I keep running into the problem where the zpool status -v
reports problems with swap filesystem and following is the depiction
of the error.
=================================================================
# zpool status -v
    pool: zpool0
   state: ONLINE
status: One or more devices has experienced an error resulting in data
          corruption.  Applications may be affected.
action: Restore the file in question if possible.  Otherwise restore
the
          entire pool from backup.
     see:http://www.sun.com/msg/ZFS-8000-8A
   scrub: none requested
          NAME        STATE     READ WRITE CKSUM
          zpool0         ONLINE       0     0     4
            c0t0d0    ONLINE       0     0     4
You have a pool without any redundancy, never a good idea.
Hey Ian,
I did not get that...pool without any redundancy? pool has two disks
with mirror configuration and this is basically for arbitrary storage
only so I do not really be using the redundant pool. (Is not the pool
by it self redundant in zfs nature?)
The status list above only shows c0t0d0.  Unlike other filesystems, ZFS
does not silently ignore errors.  If you don't have redundancy, you can
easily end up with a corrupt pool.
Setting the "copies" property to 2 may help, but if you keep getting
errors you most likely have either bad memory, or a bad drive.
Post by Nikhil
the swap is the only filesystem that runs into the problem.. so you
meant to say that zpool redundancy could be related? but how?
What other filesystems are on the pool?
--
Ian Collins- Hide quoted text -
- Show quoted text -
ah.. I forgot to mention I am sorry for the confusion here.. this is
because of the hardware raid that I am using, although it shows 1
disk, there are 2 disks on the system and they are through hardware
raid.

there are other 2 filesystems that are primarily used for generating
random and arbitrary data and can be got rid of.. so we put zpool
configuration whenever the machine netboots.
Ian Collins
2010-08-24 09:22:35 UTC
Permalink
Post by Nikhil
Post by Ian Collins
Post by Nikhil
I did not get that...pool without any redundancy? pool has two disks
with mirror configuration and this is basically for arbitrary storage
only so I do not really be using the redundant pool. (Is not the pool
by it self redundant in zfs nature?)
The status list above only shows c0t0d0. Unlike other filesystems, ZFS
does not silently ignore errors. If you don't have redundancy, you can
easily end up with a corrupt pool.
Setting the "copies" property to 2 may help, but if you keep getting
errors you most likely have either bad memory, or a bad drive.
Post by Nikhil
the swap is the only filesystem that runs into the problem.. so you
meant to say that zpool redundancy could be related? but how?
What other filesystems are on the pool?
[please trim google nonsense]
Post by Nikhil
ah.. I forgot to mention I am sorry for the confusion here.. this is
because of the hardware raid that I am using, although it shows 1
disk, there are 2 disks on the system and they are through hardware
raid.
That's generally not a good idea. If you can, let ZFS do the mirroring.
--
Ian Collins
Nikhil
2010-08-24 09:49:32 UTC
Permalink
Post by Ian Collins
Post by Nikhil
Post by Nikhil
I did not get that...pool without any redundancy? pool has two disks
with mirror configuration and this is basically for arbitrary storage
only so I do not really be using the redundant pool. (Is not the pool
by it self redundant in zfs nature?)
The status list above only shows c0t0d0.  Unlike other filesystems, ZFS
does not silently ignore errors.  If you don't have redundancy, you can
easily end up with a corrupt pool.
Setting the "copies" property to 2 may help, but if you keep getting
errors you most likely have either bad memory, or a bad drive.
Post by Nikhil
the swap is the only filesystem that runs into the problem.. so you
meant to say that zpool redundancy could be related? but how?
What other filesystems are on the pool?
[please trim google nonsense]
Post by Nikhil
ah.. I forgot to mention I am sorry for the confusion here.. this is
because of the hardware raid that I am using, although it shows 1
disk, there are 2 disks on the system and they are through hardware
raid.
That's generally not a good idea.  If you can, let ZFS do the mirroring.
--
Ian Collins- Hide quoted text -
- Show quoted text -
that was on the point that we had hardware raid perform better than
zfs mirroring performance..I do not have the statistics on hand but I
am still not able to figure out what could be the cause for this
zpool0:swap<0x1> problems? what is an easy fix without rebooting/
reconfiguring the zpool0 or zpool0/swap filesystem.

Nikhil
Andrew Gabriel
2010-08-24 14:55:09 UTC
Permalink
Post by Nikhil
Post by Ian Collins
That's generally not a good idea.  If you can, let ZFS do the mirroring.
that was on the point that we had hardware raid perform better than
zfs mirroring performance..I do not have the statistics on hand but I
am still not able to figure out what could be the cause for this
zpool0:swap<0x1> problems? what is an easy fix without rebooting/
reconfiguring the zpool0 or zpool0/swap filesystem.
If you let ZFS handle the redundancy, it will tell you what's wrong
and fix it automatically. However, you chose to give that task to your
hardware raid, so you will have to ask it what's wrong and how to fix
it. (It probably won't have a clue, which is why that was a bad choice.)
Anyway, it sounds like you deliberately decided to chose speed over data
integrity.
--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]
Continue reading on narkive:
Search results for 'zpool swap errors' (Questions and Answers)
3
replies
what is the zfs?
started 2006-11-02 01:13:32 UTC
hardware
Loading...