SOLVED Upgrade from FreeNAS 9.1 to 9.10.2; /dev/da1s2a: no such file or directory?

[stewartLG]

I've had a FreeNAS system running for a number of years that I use irregularly, often powering it down for long periods. I tried bringing it up again today, and found it would no longer properly share my CIFS share, and the Samba services looked like they were having problems/crashing. Since it's an old install of FreeNAS 9.1, I thought I would upgrade it as a first step. (I subscribe to the principle of if it ain't broke, don't fix it, but if it is broke, fix everything at once.)

The system is a old server with 6 large SATA drives, and an installation of FreeNAS 9.1 on an internally mounted, bootable USB key.

I read the Upgrading Using the ISO instructions here: http://doc.freenas.org/9.10/install.html. I made a new bootable USB key with a FreeNAS 9.10.2 on it, and booted from just that. I was offered just my 6 SATA drives as an installation candidate, and I wasn't sure that was right -- don't I want to update my existing USB key?

So I rebooted, inserting both keys this time. 9.10.2 booted, and I followed the instructions, choosing "da1" as my installation target, which now appeared in my list of drives. But then the installer crashes out with this error:

Code:

mount: /dev/da1s2a: No such file or directory.
The FreeNAS upgrade on DA1 has failed. Press Enter to continue.

I checked my dev directory, and could find only these devices:

Code:

/dev/da1
/dev/da1s1
/dev/da1s1a
/dev/da1s2
/dev/da1s3

...

I have backed up my configuration already, so I'm OK with a re-install rather than an upgrade, if that's what it takes. I can do a bit of work to get my shares/services back online. But I absolutely do not want to injure the data on the large SATA drives, which makes me skitish about trying things without certainty they'll work.

I really like FreeNAS, and have a number of boxes running it, but I've never been completely clear on if it is fundamentally a bootable live CD, or a living breathing OS. This confusion may be contributing to my confusion about what to do with this upgrade.

Thank you!

 

[Ericloewe]

What hardware are we talking about? And what USB drives are these?

I really like FreeNAS, and have a number of boxes running it, but I've never been completely clear on if it is fundamentally a bootable live CD, or a living breathing OS. This confusion may be contributing to my confusion about what to do with this upgrade.

As of 9.3, it's the latter. Boot device use is still limited, but it doesn't run from a ramdisk.

 

[stewartLG]

This an old battleship of an Opteron RAID box:

http://www.avsforum.com/forum/26-ho...ss-expensive-norco-4220-4224-alternative.html

I did not replace the power supply or fans, so it's stupidly loud, but it's 64 bit and has 16 GB of RAM.

The USB keys are nothing special, but I think that trying to install on top of an old FreeNAS installation was definitely causing issues. In the end I installed to a completely new key, loaded my old configuration, and everything came online with minimal fuss. Great!

I found one of the drives had 424 "offline uncorrectable sectors", so I'm assuming that drive is close to death and I've ordered a replacement.

Code:

CRITICAL:  Dec. 22, 2016, 11:40 a.m. - Device: /dev/ada1, 424 Currently unreadable (pending) sectors
CRITICAL:  Dec. 22, 2016, 11:40 a.m. - Device: /dev/ada1, 424 Offline uncorrectable sectors
CRITICAL:  Dec. 22, 2016, noon - Device: /dev/ada1, 440 Currently unreadable (pending) sectors
CRITICAL:  Dec. 22, 2016, noon - Device: /dev/ada1, 440 Offline uncorrectable sectors
CRITICAL:  Dec. 22, 2016, 1:01 p.m. - Device: /dev/ada1, Self-Test Log error count increased from 17 to 18
WARNING: Dec. 22, 2016, 11:40 a.m. - New feature flags are available for volume BigMediaToo. Refer to the "Upgrading a ZFS Pool" section of the User Guide for instructions.
CRITICAL:  Dec. 22, 2016, 11:40 a.m. - The capacity for the volume 'BigMediaToo' is currently at 98%, while the recommended value is below 80%.

But in the meantime, I've noticed that the available space on the ZFS volum/pool is 0 bytes, and I can't free any up. I started moving more-essential files off the drive, but the size of the array just keeps decreasing, and 0 bytes are free. For example the volume was shown as 10 TB, 0 bytes free, and now that I've moved a bunch of stuff it's 9.9 TB, 0 bytes free.

What's going on?

 

[stewartLG]

I've upgraded the ZFS pool feature flags, thinking that might explain why I was in some kind of semi-degraded/semi-read only state, but there's no change - space just keeps decreasing.

I've been running a large move operation for a while now, and files are coming off the drive properly, but the amount of free space just keeps decreasing. I'm now at 9.8 TB.

The Zpool (a ZFS2) has claimed to be healthy throughout.

Help?

Code:

[root@bauer] /mnt/BigMediaToo# zpool status -x
all pools are healthy
[root@bauer] /mnt/BigMediaToo# zpool upgrade BigMediaToo
This system supports ZFS pool feature flags.

Enabled the following features on 'BigMediaToo':
  multi_vdev_crash_dump
  spacemap_histogram
  enabled_txg
  hole_birth
  extensible_dataset
  embedded_data
  bookmarks
  filesystem_limits
  large_blocks

[root@bauer] /mnt/BigMediaToo# df -h .
Filesystem  Size  Used  Avail Capacity  Mounted on
BigMediaToo  9.8T  9.8T  0B  100%  /mnt/BigMediaToo
[root@bauer] /mnt/BigMediaToo# zpool status -x
all pools are healthy
[root@bauer] /mnt/BigMediaToo# zpool status
  pool: BigMediaToo
state: ONLINE
  scan: scrub repaired 0 in 29h8m with 0 errors on Tue Apr 26 05:08:45 2016
config:

  NAME  STATE  READ WRITE CKSUM
  BigMediaToo  ONLINE  0  0  0
  raidz2-0  ONLINE  0  0  0
  gptid/37d26165-ffd8-11e2-9dba-003048c531c4  ONLINE  0  0  0
  gptid/388574b6-ffd8-11e2-9dba-003048c531c4  ONLINE  0  0  0
  gptid/393e0631-ffd8-11e2-9dba-003048c531c4  ONLINE  0  0  0
  gptid/39edbf4e-ffd8-11e2-9dba-003048c531c4  ONLINE  0  0  0
  gptid/3aa330ff-ffd8-11e2-9dba-003048c531c4  ONLINE  0  0  0
  gptid/3b5cceb0-ffd8-11e2-9dba-003048c531c4  ONLINE  0  0  0

errors: No known data errors

  pool: freenas-boot
state: ONLINE
  scan: none requested
config:

  NAME  STATE  READ WRITE CKSUM
  freenas-boot  ONLINE  0  0  0
  da0p2  ONLINE  0  0  0

errors: No known data errors
[root@bauer] /mnt/BigMediaToo#

And a few minutes later..

Code:

[root@bauer] /mnt/BigMediaToo# df -h .
Filesystem  Size  Used  Avail Capacity  Mounted on
BigMediaToo  9.7T  9.7T  0B  100%  /mnt/BigMediaToo

And here are some other random commands. I am not sure what I'm looking for, perhaps this will help someone diagnose what's going on? If not, tell me what you need.

Code:

[root@bauer] /mnt/BigMediaToo# zpool list
NAME  SIZE  ALLOC  FREE  EXPANDSZ  FRAG  CAP  DEDUP  HEALTH  ALTROOT
BigMediaToo  21.8T  21.3T  424G  -  5%  98%  1.00x  ONLINE  /mnt
freenas-boot  3.72G  652M  3.08G  -  -  17%  1.00x  ONLINE  -

Code:

[root@bauer] /mnt/BigMediaToo# zfs list
NAME  USED  AVAIL  REFER  MOUNTPOINT
BigMediaToo  14.2T  0  9.75T  /mnt/BigMediaToo
BigMediaToo/JailsDataset  1.61G  0  69.8M  /mnt/BigMediaToo/JailsDataset
BigMediaToo/JailsDataset/.warden-template-9.1-RELEASE-amd64  429M  0  429M  /mnt/BigMediaToo/JailsDataset/.warden-template-9.1-RELEASE-amd64
BigMediaToo/JailsDataset/NagiosJail  1.13G  0  1.54G  /mnt/BigMediaToo/JailsDataset/NagiosJail
freenas-boot  652M  2.96G  64K  none
freenas-boot/ROOT  645M  2.96G  29K  none
freenas-boot/ROOT/Initial-Install  1K  2.96G  635M  legacy
freenas-boot/ROOT/default  645M  2.96G  638M  legacy
freenas-boot/grub  6.35M  2.96G  6.35M  legacy

 

[stewartLG]

Aha!

It turns out I had a huge number of old snapshots. When I deleted them, the space was freed up.

My server is now working fine, although I will be replacing the failing drive soon.

I'm unsure how to mark the thread Solved, can I do that myself?

 

[Ericloewe]

Yes you can. There's an "edit thread" link near the top of the page.

 

[stewartLG]

When I search for "Edit Thread" on this page, the only mention is your post.

 

[wblock]

Thread Tools -> Edit Thread.