Rebooting while copying files from a pool to a disk

[pettazz]

I'm trying to copy everything off of my existing, RAID-0 pool onto a separate disk so that I can start putting together a better setup than the mess I have now. I have an 8TB HDD and ~4.5TB of stuff to save onto it. I connected the 8TB internally via SATA, partitioned and formatted it with ext2 (to avoid dealing with any fuse filesystems) and mounted it to /stuff-backup. I currently have one Volume: /mnt/stuff, made up of 5 disks in a stripe (I know, I know, this is why I'm redoing it), and with a few datasets: /mnt/stuff/Media, /mnt/stuff/Backups, etc. Every time I try to copy files from the volume to the HDD, my system hangs up for a few minutes and then reboots. I've tried:

cp -Rv /mnt/stuff/ /stuff-backup/
cp -Rv /mnt/stuff/Backups/ /stuff-backup/Backups/
cp -Rv /mnt/stuff/Backups/some-specific-dir /stuff-backup/Backups/
rsync -vrzh /mnt/stuff/ /stuff-backup/
rsync -vrzh /mnt/stuff/Media/ /stuff-backup/Media/

All to the same effect: they start working for a few minutes, then the system reboots.

/var/crash is empty, and /var/log/messages doesn't seem to have anything relevant, just normal stuff then suddenly booting:

Code:

Mar 29 11:07:57 wilf afpd[6604]: AFP statistics: 644142.03 KB read, 347220.41 KB written
Mar 29 11:07:57 wilf afpd[6604]: done
Mar 29 11:10:55 wilf syslog-ng[1560]: syslog-ng starting up; version='3.6.4'
Mar 29 11:10:55 wilf Copyright (c) 1992-2016 The FreeBSD Project.

and I haven't seen anything else seemingly interesting elsewhere in /var/log/. last shows me a bunch of boot time messages but without any crash or shutdown time messages before them. Looking at the reporting graphs, I don't see Memory, CPU usage, or Temperatures getting out of hand. My power supply is more than enough to handle the hardware in this machine twice over.

So my questions are:

• Am I trying to do something the wrong way here? Is it not acceptable to try to copy files directly from the mount points, and I should just go mount this disk on another machine and transfer (much more slowly) over the network?

or

• Is there somewhere else I can look for more details on what's causing these reboots? I'm coming from more of a Linux background, so I don't have a lot of experience with FreeBSD. I've seen mentions while googling of a motherboard watchdog rebooting hung systems, but mine is disabled in BIOS, and I don't see why just copying files would hang up the whole system every time.

 

[joeschmuck]

Post the specs of your system and what version of FreeNAS you are running. You might be better off taking that 8TB drive and just creating a single striped drive and then try to copy the files to it.

 

[pettazz]

Added some details to my signature.

If I do the single stripe maneuver, once I create my new pool, doing "Import Volume" will just copy all the data into the pool and free up the disk (like Import Disk)?

 

[joeschmuck]

No, the data will not automatically copy over. When you create the new volume, it's just a new drive on the system, meaning if you call it "savemystuff" then it will be /mnt/savemystuff/ and you can then copy the data via the command line like you were trying to do earlier. After you have copied all your data over, I'd power off the machine and disconnect the cable(s) from that 8TB drive, power on the machine, blow away your current pool, recreate your pool the way you want it, power off, reconnect the 8TB drive, power on, and lastly, copy all that data back via the command line. Once all is done you can remove (do not destroy) the 8TB drive, power off, remove the drive, power on and verify all is okay. Or you could leave that 8TB drive installed and use it somehow, maybe for non-important data but you assume the risk of data loss as if that drive fails, the data on uit would be gone.

Simple! Right?

 

[pettazz]

Oh I see what you're saying. If I still run into this rebooting issue copying it that way I think it might be time to consider looking into the warranty on this motherboard.

I'll give it a try, thanks.

 

[gpsguy]

Added some details to my signature.
?

In the future, if you ask for help, please put the detailed info in the body of your message. Those of us on mobile devices can't see your signature. TIA

 

[snaptec]

I can't see your sig as I'm mobile.
Maybe your psu is faulty?
If it's stressed from all drives and a bit CPU in action a crash can happen.
You won't have any logs if that happens.
Do you have IPMI?
Maybe look at the hw stats while start copying.


Gesendet von iPhone mit Tapatalk

 

[pettazz]

Ah, didn't know it didn't work on mobile. This is my guy:

FreeNAS-9.10.2-U2 | ASRock MiniITX C2550D4I | Intel Avoton C2550 Quad-Core 2.41GHz | 16 GB DDR3 ECC RAM | A bunch of old garbage SATA2 HDDs | 2x 8TB WD Red

joeschmuck's idea seems to be working! It's been copying from my old pool onto the 8TB in a single stripe volume for about two hours now with no trouble. I'm wondering if this is somehow a problem with ext2fs, or at least with copying between zfs/ext2fs. Once the copy is (knock on wood) completed, I'll try digging around for some logs about it.

 

[joeschmuck]

You guys with smart phones are just too advanced for me. I still like my flip phone and can get away with charging it only once a week, twice a week when I'm talking a bit more. ;)