panic...

[gdcmjb]

Hi,

I'm using freenas for a long time now when it stop working on its own...

Here's the message I got :

Code:

Mounting local file systems:.
panic: Solaris(panic): zfs: freeing free segment (offset=9556128265949675776 size=7020108323661563922)

cupid = 0
KDB: enter: panic
[thread pid 0 tid 100083 ]
Stopped at kdb_enter+0x3b: movl $0,kdb_why

Can someone help me?

After several test, it seams that if i disconnect the 2 first hdd (4 total) the system starts well...

THX in advance !

 

[gdcmjb]

Can someone help me ? I'm really scared loosing all my docs and pics saved on nas...

Thx :)

 

[William Grzybowski]

How much RAM have you got?

Paste output of # zpool status

 

[gdcmjb]

How much RAM have you got?

Paste output of # zpool status

Hi, thx for the reply

1Gb, what's # zpool status ?

 

[peterh]

i's suggest 2 things to do in order :

1/ read the documentation

2/ install at least 4G om memory and use a 64-bit system ( all of this is mentioned in the documentation)

if you don't upgrade memory, kiss your data goodbye and start using ufs instead.

 

[gdcmjb]

i's suggest 2 things to do in order :

1/ read the documentation

2/ install at least 4G om memory and use a 64-bit system ( all of this is mentioned in the documentation)

if you don't upgrade memory, kiss your data goodbye and start using ufs instead.

Hi, thx for reply.

I've updated to 8G ram, still the same error message.

Got an idea?

Thx

 

[gdcmjb]

Sorry, not really the same message, there're some little differences (in bold) :

Code:

Mounting local file systems:.
panic: Solaris(panic): zfs: freeing free segment (offset=9771122371598483712 size=7020108323711624896)

cupid = 2
KDB: enter: panic
[thread pid 0 tid 100199 ]
Stopped at kdb_enter+0x3b: movl $0,kdb_why

 

[William Grzybowski]

What freenas version are you running

 

[gdcmjb]

What freenas version are you running

Hi,

8.2

Thx

 

[gdcmjb]

Paste output of # zpool status

I tried

Code:

zpool status -v

and I got

Code:

No such command

 

[William Grzybowski]

You need to run as root...

Do you have autotune enabled? (System->Advanced)
Do you have tunables? (System->Tunables)

 

[gdcmjb]

You need to run as root...

Do you have autotune enabled? (System->Advanced)
Do you have tunables? (System->Tunables)

Where can I find this in command line? The system do not start, it just stop after the previous error message and I'm only able to type something after

Code:

db >

 

[William Grzybowski]

Ah, so the system does not even boot...

Try download FreeBSD 9.1 RC2 (freebsd.org), run the luve usb image, go to shell and run:

# zpool import POOLNAME

FreeBSD has a newer ZFS version that might fix the issue.

 

[gdcmjb]

allright, i'll try this

Thx !

 

[gdcmjb]

Ah, so the system does not even boot...

Try download FreeBSD 9.1 RC2 (freebsd.org), run the luve usb image, go to shell and run:

# zpool import POOLNAME

FreeBSD has a newer ZFS version that might fix the issue.

I did the trick, and I got this :

Code:

panic: Solaris(panic): zfs: freeing free segment (offset=******XX size=*********XX)
cpuid = 2
KDB: stack backtrace:
#0 0xc0af3aff at kdb_backtrace+0x4f
#1 0xc0ac052ff at panic+0x16f
#2 0xca669c71 at vcmn_err+0x71
#3 0xca58eead at zfs_panic_recover+0x2d
#4 0xca591ea2 at space_map_remove+0x302
#5 0xca592835 at space_map_load+0x275
#6 0xca578369 at metaslab_activate+0x139
#7 0xca579712 at metaslab_alloc+0xc12
#8 0xca5c0d22 at zio_dvs_allocate+0xa2
#9 0xca5bd101 at zio_execute+0xa1
#10 0xca5bd32f at zio_notify_parent+0xff
#11 0xca5bf53b at zio_ready+0xeb
#12 0xca5bd101 at zio_execute+0xa1
#13 0xc0affee9 at taskqueue_run_locked+0xc9
#14 0xc0b00ddc at taskqueue_thread_loop+0xbc
#15 0xc0a90526 at fork_exit+0x96
#16 0xc0e0f6e4 at fork_trampoline+0x8

(wouha.... it was long to type :))

 

[gdcmjb]

I don't know if it's important but to get this I had to type

Code:

zpool import -f GDNAS-01

With

Code:

zpool import GDNAS-01

I get

Code:

cannot import 'GDNAS-01': pool may be in use from other system, it was last accessed by (hostid: 0xe4001ae7) on Wed Sep 26 10:34:39 2012 use '-f' to import anyway

For your info Wed Sep 26 10:34:39 2012 is the date of the crash...

 

[William Grzybowski]

Sorry, I don't know what else you could try... Looks like your pool corrupted to some kind of unrecoverable state.

Maybe someone else has an idea.

 

[gdcmjb]

Sorry, I don't know what else you could try... Looks like your pool corrupted to some kind of unrecoverable state.

Maybe someone else has an idea.

... :( I hope that someone'll have an idea... I just CAN'T loose years of memories pictures, specially my little one year old girl's pics from her birth...

Anyway thx for help ;)

 

[William Grzybowski]

Ok, I got one more idea left.

Boot FreeBSD 9.1, on the boot loader choose prompt (option 6 IIRC)

type:

set vfs.zfs.recover=1 <enter>
boot <enter>

Once the OS boots, enter this in the shell:
zpool import -F -f -o readonly=on POOLNAME

If you're lucky this will at least give you RO access to the pool.

 

[Stephens]

And then remember... RAID is not backup. Backup your important stuff somewhere other than your NAS.