Replace old disk by a new one. (Move pool to new disk)

[Ismael Duarte]

Hi guys,
I've read some post and found some information, but I'm afraid to make some mistake and loose all or most of my data. So please give me an help.

My FreeNas version is: FreeNAS-8.3.1-RELEASE-x86 (r13452)
At this moment my pool is:

Code:

zpool list DISCO
NAME    SIZE  ALLOC   FREE    CAP  DEDUP  HEALTH  ALTROOT
DISCO  1.81T  1.34T   481G    74%  1.00x  ONLINE  /mnt

Code:

zpool status
  pool: DISCO
 state: ONLINE
  scan: scrub repaired 0 in 0h0m with 0 errors on Sun Mar 24 11:20:18 2013
config:

        NAME                                          STATE     READ WRITE CKSUM
        DISCO                                         ONLINE       0     0     0
          gptid/b508afe4-2175-11e1-9bde-0013d4190f38  ONLINE       0     0     0
          gptid/0e99edd8-21e6-11e1-8c42-0013d4190f38  ONLINE       0     0     0
          gptid/825ab845-6235-11e2-824e-0013d4190f38  ONLINE       0     0     0

I bought a new hardrive of 2TB, I would like to copy all data to the new HDD and at the end have the same pool name, so all shares remain intact. The reason for this is one of the disk of the actual pool is showing some signals to colapse.

Appreciate your help

 

[Ismael Duarte]

Re: Replace old disk bya a new one.

Hi again,
I've tryed to start the work, while I'm waiting some help. Tryed to import the new disk as UFS, but .. there's a problem at this first step.
Can't import this this, as I've done before :( always receive the mesage "The selected disks were not verified for this import rules."
What can I do?

 

[ProtoSD]

Re: Replace old disk bya a new one.

http://doc.freenas.org/index.php/Volumes#Replacing_a_Failed_Drive_or_Zil_Device

 

[Ismael Duarte]

Re: Replace old disk bya a new one.

http://doc.freenas.org/index.php/Volumes#Replacing_a_Failed_Drive_or_Zil_Device

Thank you ProtoSD
I think this is a little diferent!
I've a pool with 3 disk with the capacity of 1,8TB
I bought a new HDD and want to copy all data to new disk e remove the old ones.

I've managed to import the new 2TB HDD to FreeNas with a new pool name. Now I'll start to copy all data to the new pool.
I hope I'm doing right. Than I think I can disconnect the drives from old pool.
After that I need to find out how to rename the new pool to DISCO, the name of the old pool (help here?)
hopefully I'll have everything working ok in the new hdd

Thank you once more.

 

[Caesar]

Re: Replace old disk bya a new one.

I believe the docs cover renaming a pool.... I am curious tho why would you go from a 3 drive setup (assuming there's redundancy) to a one drive setup without any redundancy?

 

[gpsguy]

Re: Replace old disk bya a new one.

At first, I *thought* the same thing.

Based on the results of "zpool status" it doesn't appear that he has any redundancy. Though RAID<>Backup, I hope he has backups of any critical data.

I am curious tho why would you go from a 3 drive setup (assuming there's redundancy) to a one drive setup without any redundancy?

 

[Caesar]

Re: Replace old disk bya a new one.

At first, I *thought* the same thing.

Based on the results of "zpool status" it doesn't appear that he has any redundancy. Though RAID<>Backup, I hope he has backups of any critical data.

Thanks gpsguy.... I am not that familiar with zpool status so i missed that. I checked mine and it does show raidz1-0 so I assume it would say raidz2-0 if I had raidz2

 

[cyberjock]

Re: Replace old disk bya a new one.

Thanks gpsguy.... I am not that familiar with zpool status so i missed that. I checked mine and it does show raidz1-0 so I assume it would say raidz2-0 if I had raidz2

Correct.

 

[Ismael Duarte]

No redundancy and important data :(
But, can't afford more disks
I've a problem and I'm affraid for my data. I left freenas over night copying data. In the morning I've received a Segmentation Fault and after reboot freenas doesn't work.
At the afternoon I'l try to figure out what's wrong, hope you can help me.

 

[cyberjock]

No redundancy and important data :(
But, can't afford more disks.

That's a really bad combo. This is why I don't recommend anything less than RAIDZ2, especially to people that choose not to do religious backups.

If you can't afford more disks this usually ends with "then you can obviously afford to replace your data when its lost". As much as that sucks, it seems to be a fundamental truth more often then not for people that don't do regular backups. :(

I always tell people that your data is only worth what you are willing to spend to protect it. Don't want to spend any money on good hardware and backups, then your data must be worthless. That's what you valued it at. Spent $2000 on good solid backup hardware? Then came up with a good solid backup procedure? Then that's what your data is worth.

 

[Ismael Duarte]

Agree with you. I know that but at this momment there's nothing more I can do.
The only thing I hope is this community to help me when needed as much as possible.

As I said before, I left the Nas copying data overnight but as stoped, at CLI I've got only "Segmentation Fault" as answer.
Now is working again.
What's de best way to copy data between pools? I've done cp -a DISCO/* DATA/
DISCO = Old Pool, DATA = new pool.

 

[paleoN]

As I said before, I left the Nas copying data overnight but as stoped, at CLI I've got only "Segmentation Fault" as answer.

So, cp seg faulted? That's unusual. Have you been having any other problems with the NAS?

What's de best way to copy data between pools?

Either zfs send/receive or rsync.

 

[ProtoSD]

I'd run a memtest to check your RAM, if that turns out ok, then do a scrub to make sure your pool didn't get corrupted. Don't do the scrub before you check your RAM. If the RAM is bad, replace it before doing a scrub.

 

[cyberjock]

I wouldn't do the scrub with your disk missing.

 

[Ismael Duarte]

Thank you all guys.
There's no disk missing. RAM was testes some months ago and scrub. I think the problem was (I Hope) not enough RAM for the new drive.
I've bought some more, I've now 4GB (was 2,5). I'm trying again.
How that command work? zfs send/receive?
I've tried with pax also, same result. Let's see now with more RAM.

Thank you once more.

------

Don't understand :( a lot of errors copying from /mnt/DISCO to /mnt/DATA
After some time I've a system crash and must power off/on

Is there any better way to do it?

 

[Ismael Duarte]

After some of this ... Segm. fault ...

Code:

cp: failed to set acl entries on DATA/mydata: Operation not supported

Another clue, received this by email from root:

Code:

newsyslog: log_trim: fclose: File too large

Code:

Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: vm_fault: pager read error, pid 5570 (cron)
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: pid 5570 (cron), uid 0: exited on signal 11
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: vnode_pager_getpages: I/O read error
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: vm_fault: pager read error, pid 5574 (cron)
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: pid vnode_pa5g5e7r4_ g(ectrpoang)e,s :u iId/ O0 :r eeaxidt eedr roonr signal 11
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: 
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: vm_fault: pager read error, pid 5572 (cron)
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: pid 5572 (cron), uid 0: exited on signal 11
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: vnode_pager_getpages: I/O read error
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: vm_fault: pager read error, pid 5575 (cron)
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: vnode_pag
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: e<r_6g>etppiad ges: I/O re5a5d7 5e r(rcorron), uid 0: exited on signal 11
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: 
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: vm_fault: pager read error, pid 5573 (cron)
Mar 27 19:59:58 192.168.1.3 Mar 27 20:00:00 kernel: pid 5573 (cron), uid 0: exited on signal 11
Mar 27 20:00:58 192.168.1.3 Mar 27 20:01:00 kernel: vnode_pager_getpages: I/O read error
Mar 27 20:00:58 192.168.1.3 Mar 27 20:01:00 kernel: vm_fault: pager read error, pid 5586 (cron)
Mar 27 20:00:58 192.168.1.3 Mar 27 20:01:00 kernel: pid 5586 (cron), uid 0: exited on signal 11
Mar 27 20:01:58 192.168.1.3 Mar 27 20:02:01 kernel: vnode_pager_getpages: I/O read error
Mar 27 20:01:58 192.168.1.3 Mar 27 20:02:01 kernel: vm_fault: pager read error, pid 5596 (cron)
Mar 27 20:01:58 192.168.1.3 Mar 27 20:02:01 kernel: pid 5596 (cron), uid 0: exited on signal 11
Mar 27 20:02:58 192.168.1.3 Mar 27 20:03:01 kernel: vnode_pager_getpages: I/O read error
Mar 27 20:02:58 192.168.1.3 Mar 27 20:03:01 kernel: vm_fault: pager read error, pid 5606 (cron)
Mar 27 20:02:58 192.168.1.3 Mar 27 20:03:01 kernel: pid 5606 (cron), uid 0: exited on signal 11
Mar 27 20:03:58 192.168.1.3 Mar 27 20:04:00 kernel: vnode_pager_getpages: I/O read error
Mar 27 20:03:58 192.168.1.3 Mar 27 20:04:00 kernel: vm_fault: pager read error, pid 5616 (cron)
Mar 27 20:03:58 192.168.1.3 Mar 27 20:04:00 kernel: pid 5616 (cron), uid 0: exited on signal 11
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: vnode_pager_getpages: I/O read error
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: vm_fault: pager read error, pid 5626 (cron)
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: vnode_pager_getpages: I/O read error
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: vm_fault: pager read error, pid 5628 (cron)
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: pid 5626 (cron), uid 0: exited on signal 11
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: vnode_pager_getpages: I/O read error
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: vm_fault: pager read error, pid 5629 (cron)
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: pid 5628 (cron), uid 0: exited on signal 11
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: pid 5629 (cron), uid 0: exited on signal 11
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: vnode_pager_getpages: I/O read error
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: vm_fault: pager read error, pid 5627 (cron)
Mar 27 20:04:58 192.168.1.3 Mar 27 20:05:00 kernel: pid 5627 (cron), uid 0: exited on signal 11

 

[Ismael Duarte]

Hey Guys, a little help here, please.
I'm still fighting to recover my data, at least some.
I'm running freenas with same disks on my main computer. A 2.8Ghz quad-core with 7GB RAM
Every-time I try to import pool DISCO the system halts with TRAP12 (DISCO is where I have all data)
RAM is good, was tested once more.

Is all data lost for ever? What else can I try.

Thank you!

 

[cyberjock]

In the first post you said you had version FreeNAS-8.3.1-RELEASE-x86 (r13452). Is it not the x64 version? You have more than 4GB of RAM, so you should be using the 64 bit version. FreeNAS can't use more than 4GB with the x86 version.

If you really are using the x86 version I would try making a new USB stick with the x64 version, import your zpool, and try to use the system again. By using the x86 version of FreeNAS you are limiting the system to 4GB(technically probably 3.25GB or so).

Keep in mind that each time your system isn't shut down properly and you are forced to hit the reset button or the power button you are putting your zpool at risk because of possible incomplete writes. This is why it's stressed so much to have enough RAM for ZFS or to use UFS. An unstable system can cause data loss.

 

[Ismael Duarte]

cyberjock, thank you for replying me.
FreeNAS-8.3.1-RELEASE-x86 (r13452) is my system. But, when trying to solve this problem, I've done what you said here. I've used my main computer with a new USB Stick with FreeNAS x64. Didn't solve, when trying to import DISCO pool it always get blocked with FATAL TRAP 12.

So my big question now is, Is there any other way to recover my data, por part of it?

Thank you for your time.

In the first post you said you had version FreeNAS-8.3.1-RELEASE-x86 (r13452). Is it not the x64 version? You have more than 4GB of RAM, so you should be using the 64 bit version. FreeNAS can't use more than 4GB with the x86 version.

If you really are using the x86 version I would try making a new USB stick with the x64 version, import your zpool, and try to use the system again. By using the x86 version of FreeNAS you are limiting the system to 4GB(technically probably 3.25GB or so).

Keep in mind that each time your system isn't shut down properly and you are forced to hit the reset button or the power button you are putting your zpool at risk because of possible incomplete writes. This is why it's stressed so much to have enough RAM for ZFS or to use UFS. An unstable system can cause data loss.

 

[cyberjock]

My guess is because you were running with <4GB of RAM(because you were using x86) and the excessive manual reboots you have had to do you aren't likely to get your data back.

You could read through this thread and try a few of these commands... http://forums.freenas.org/showthrea...uddenly-voluime-storage-(ZFS)-status-unknown-! Keep in mind anything that forces a mount as a parameter can be dangerous and shouldn't necessarily be done until you've exhausted the other options.

I don't have alot of time today to sit down and try to help you restore your data unfortunately, but any commands I recommend are likely very similar or identical to those in that thread.