NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
Solved
PRO6 "Remove inactive volumes" after interrupted balance
Firmware 6.9.3, RAID 6, 6 drives. I let the volume get really full before my autoscheduled balance kicked on and ground everything to a halt. Forced reboot gives all red drives and the "Remove inac...
Hi itachi2,
It appears that the volume is already filled up when the balance ran which has given it no room to rewrite the data. When the file system comes up, it is trying to finish a large amount of previous tasks but there is no space so it just locks up.
You can try booting the NAS to read-only and check if you can access the data that way. Otherwise you have to look for the 3rd party data recovery to assist you in recovering the data.
As what StephenB mentioned, Pro 6 with OS6 firmware is not supported by NETGEAR.
Regards,
I can't be 100% sure, but I think my last reply to this thread may have been deleted. Prior to this message I have 3 posts, 2 of which are replies. Those two probably are the missing message in question and my edit of that message for grammar and style. In this message I thanked both Stephen and John for their help and suggested other ways to help that could potentially assist users who either can't afford or are not eligible for support.
I also mentioned a few btrfs and fsck commands as potential avenues for the user to self-repair, and suggested that with such a high amount of experience and wisdom regarding Linux and btrfs that it would be even more useful to nudge or guide people in a relatively safe direction to attempt to get the original data online again rather than backup and restore for a few days (even at GB speeds). I opined that many disclaimers could be given to absolve the adviser, forum and company from any ill effects that may arise from the use or misuse of the potentially proffered advice.
I mean no disrespect or insult with either this description of my mystery post or that post itself. The help I was given was on a purely volunteer and altruistic level to begin with and I am appreciative. The advice to mount in read-only is solid, and if luck is with you and time is no issue, that would allow most people in similar situations to recover their data. However, the hobbyist in me can only think, "This doesn't _fix_ the problem that happened." Though specifics would be unique to each volume, there is a methodical flowchart that one could follow depending on the error messages and output from commands given.
There have been participants in threads in the past who did go above and beyond to share the benefit of their breadth of knowledge to the folks in need. Maybe I'm naive, but I don't see a problem with even a company rep suggesting a particular mount, fsck or btrfs command with the imperative to make sure that data is backed up first and the usual disclaimers regarding said commands and data loss and liability, etc.
And I also understand that this is in "Using your ReadyNAS," not "Data Recovery 101 for the Home Hacker," but I would argue that even with good use practices, sometimes it is useful, necessary, and edifying to go under the hood to tinker with the guts of the OS and filesystem, and that knowledge guided by wisdom will never be a bad thing.
So I'll mark the answer given as a solution, but not the solution I was hoping for, and it is a safe solution that uses days of time when an hour or two with well-crafted commands could potentially repair the damaged filesystem.
In any case, my upgraded Pro 6 has been my favorite NAS to use, and I look forward to restoring my data to continue using it. Thanks again to all the forum writers and company reps for doing what they can.
itachi2 wrote:
I can't be 100% sure, but I think my last reply to this thread may have been deleted.
...
Though specifics would be unique to each volume, there is a methodical flowchart that one could follow depending on the error messages and output from commands given.
It was just caught by the spam filter. Mods generally check the quarantine manually and release false positives, but there sometimes is too much spam for that to be practical.
I do take your point on the value of a methodical flowchart to troubleshoot the mount failure problem. It's not a problem I've had first hand, and unfortunately I haven't seen enough posted on the solutions to feel comfortable providing specific steps.
The missing post was:
itachi2 wrote:
John and Stephen,
Thank you both for checking my logs and the input regarding my issue. Like I said, I was able to mount the md127 and md0 in recovery with seemingly no ill effects. I thought about deleting some things from my data volume in recovery and rebooting to see if it could complete the boot, but will only attempt that after I ensure a good backup to external drives, as my original plan was to backup then erase the old / create a new Data volume anyway, so nothing to lose by mucking with the drive at that point.
Currently copying everything across the network to some USB 3.0 CIFS shares mounted on my Windows 7 box, which is much faster than the ~25MB/s afforded by USB 2.0. And it seems to be going okay - I should be able to get the bulk of the data off soon.
If you were inclined, I would have liked to have heard the benefit of your collective experience with some other possibilities for further self-help and repair... i.e. "If it were my NAS, I might attempt a mount with these options: x, y, and z" or "You may want to explore the commands 'btrfs-zero-log' or 'btrfs rescue' or 'btrfsck --repair [--init-extent-tree]' or 'btrfs check --repair' but I, the forum and Netgear are not responsible for any loss or damage that may occur from the use of these commands..." The suggestion for booting in read-only is a good one, and I may have completely overlooked that when I was exploring my rescue options. I wound up mounting /data as read-only anyway in recovery for safety's sake.
Well, I wound up rambling a bit... Thanks again for the assist.
Stephen,
Thanks for reading. I wasn't necessarily talking about a literal flowchart, but more of a general branching process that one can use depending on the errors/problems and success or failure of each remediating step. The problem-solving techniques that come from research, practice, and experience.
In any case, after I backed up most everything using a readonly mount and some Windows share destinations with CIFS, I decided to try a few commands to test.
btrfs check /dev/md/Data-0 btrfs check --repair /dev/md/Data-0 mount -t btrfs -o skip_balance /dev/md/Data-0 /mnt/data btrfs balance cancel /mnt/data
The first "check" command is optional but I wanted to see the number and types of errors without committing changes. Thankfully it seemed to be just one: "root 259 inode 1753733 errors 400, nbytes wrong"
Relevant pages:
https://btrfs.wiki.kernel.org/index.php/Btrfsck
https://btrfs.wiki.kernel.org/index.php/Problem_FAQ (which says that "such errors should be fixable with 'btrfs check --repair'")
https://btrfs.wiki.kernel.org/index.php/Manpage/btrfs%285%29#MOUNT_OPTIONS
"Mount Options" directly above for the commands to mount without resuming the balance and to cancel the balance operation in progress.
Anyway, the check threw up lots of errors regarding qgroup counts and extent buffer leaks but did give me the all-important "err is 0"
Also note that the btrfsck page above mentions "check --repair" as a *last resort* after 5 preliminary recovery methods, and even then with a recommended btrfs-tools > 4.0. You should probably use the booted firmware's version of the tools over ssh vs. the altered busybox telnet shell available through the boot menu as that is more likely to have an updated version.
In any case, I think I have my original file system back, though I can't be completely sure all files are whole--everything seems to be acting fine, though... or at least as badly as I had configured it before all this. I had 60GB of space left, but I am hoping to leave a bit more for the system to use prior to the next balance - I already backed up a TB of archive data that doesn't need to stay live and will clear that space up.
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
