Hi,Yesterday I received a notification that one of the disks in my ReadyNAS NVX was dead and required replacement.I had a spare disk of the same kind (it is the 3TB Hitachi that is listed in the HCL), so I performed a replacement (with the unit on).I got two more notifications. First "Disk removal detected. [Disk 3]" and then "Data volume will be rebuilt with disk 3".After this, nothing happens. The volume is not being rebuilt and both Frontview and RAIDar report disk 3 as dead. I left if for some time after reading that it sometimes needs a few minutes to start the rebuild, but nothing happened. I had also read that sometimes rebooting the unit is necessary, which I did, but still the unit is marked as dead and rebuild does not start (I still get the "Data volume will be rebuilt with disk 3" message).The disk that was reported dead actually seems to be just fine. I tried it on my PC, created new file systems and everything works fine. A long smart test did not reveal any issues either. I then tried to re-insert the original disk (after removing all partitions) and same thing happens. I get the notifications that the disk was removed and that a the data volume will be rebuilt.Left it overnight, but nothing happened. It did not rebuild.The data on the unit is still accessible, so that is not the problem, but I have no redundancy.I tried many other things:- Removed disk 3, restarted the unit. Frontview reports disk 3 as dead (with the capacity and model number), when it should be “missing” or “not installed”- Inserted replacement disk. Shut down unit, powered it on again.- This is the output of fdisk (through an ssh session on the unit)~# fdisk -l /dev/sdcfdisk: device has more than 2^32 sectors, can't use all of themFound valid GPT with protective MBR; using GPTDisk /dev/sdc: 4294967295 sectors, 2047GLogical sector size: 512Disk identifier (GUID): 8b974a2e-8823-42a2-ba93-113c35cb58fdPartition table holds up to 128 entriesFirst usable sector is 34, last usable sector is 5860533134Number Start (sector) End (sector) Size Code NameIt basically reports no partitions- dmesg reportssdc: unknown partition table sdc: unknown partition table sdc: unknown partition table sdc:- I even tried putting in another disk, that I know works fine, but with a smaller capacity (1TB). It was detected by the system, I got a notification that the disk was too small for the current volume, but Frontview still reported the disk as dead, but with the old capacity and model number.I have had the unit for about 2 years and have performed expansion from 1TB to 2TB and later 3TB for each of the 4 drives. Never had any issues and I also replaced 2 failed disks during this time and that also worked fine.Any clues on what might be the problem?Is there a way to force the unit to mark the disk as not present/missing?Thanks.

You are running 4.2.16 or later so your disks would use GPT not MBR. fdisk is for use with the MBR partitioning scheme only.So do e.g.MDGM-NAS:~# sgdisk -p /dev/sdcWhat brand and model disks are these?

Hi,The disks are the Hitachi Deskstar HDS723030ALA640. As I mentioned, I have had 4 of them running for quite some time and the replacement disk I inserted is exactly the same.You are right about fdisk (which is actually part of busybox), but it still reports the partitions. If I run it for sd[abd] I can see 5 partitions. In any case, this is the output of sgdisk:~# sgdisk -p /dev/sdcDisk /dev/sdc: 5860533168 sectors, 2.7 TiBLogical sector size: 512 bytesDisk identifier (GUID): 74A6315F-A55E-4CE3-8177-6991CFEC5D96Partition table holds up to 128 entriesFirst usable sector is 34, last usable sector is 5860533134Partitions will be aligned on 2048-sector boundariesTotal free space is 5860533101 sectors (2.7 TiB)Number Start (sector) End (sector) Size Code NameThe same command for sda shows ~# sgdisk -p /dev/sdaDisk /dev/sda: 5860533168 sectors, 2.7 TiBLogical sector size: 512 bytesDisk identifier (GUID): AE4A83A1-B2AB-40E3-9FA5-6407FB062EABPartition table holds up to 128 entriesFirst usable sector is 34, last usable sector is 5860533134Partitions will be aligned on 8-sector boundariesTotal free space is 4145 sectors (2.0 MiB)Number Start (sector) End (sector) Size Code Name 1 64 8388671 4.0 GiB FD00 2 8388672 9437247 512.0 MiB FD00 5 9437256 1953520131 927.0 GiB FD00 6 1953520136 3907024064 931.5 GiB FD00 7 3907024072 5860529038 931.5 GiB FD00

If you do e.g.# hdparm -I /dev/sdcdoes it report the correct info for the disk you have inserted?

Yes, the serial number seems to be correct. If I compare (diff) the output for sda and sdc, I see only the expected differences in Serial number, device identifier and unique ID plus "DMA Setup Auto-Activate optimization", that has an asterisk to its left on the working disk, but not on the replacement one.

What about# smartctl --all /dev/sdcDoes that report any issues with the disk?

Forum Discussion

rsilva1

Aspirant

Nov 11, 2012

ReadyNAS NVX problems replacing disk

Hi,

Yesterday I received a notification that one of the disks in my ReadyNAS NVX was dead and required replacement.
I had a spare disk of the same kind (it is the 3TB Hitachi that is listed in the HCL), so I performed a replacement (with the unit on).
I got two more notifications. First "Disk removal detected. [Disk 3]" and then "Data volume will be rebuilt with disk 3".

After this, nothing happens. The volume is not being rebuilt and both Frontview and RAIDar report disk 3 as dead. I left if for some time after reading that it sometimes needs a few minutes to start the rebuild, but nothing happened.

I had also read that sometimes rebooting the unit is necessary, which I did, but still the unit is marked as dead and rebuild does not start (I still get the "Data volume will be rebuilt with disk 3" message).

The disk that was reported dead actually seems to be just fine. I tried it on my PC, created new file systems and everything works fine. A long smart test did not reveal any issues either.

I then tried to re-insert the original disk (after removing all partitions) and same thing happens. I get the notifications that the disk was removed and that a the data volume will be rebuilt.
Left it overnight, but nothing happened. It did not rebuild.
The data on the unit is still accessible, so that is not the problem, but I have no redundancy.

I tried many other things:

fdisk: device has more than 2^32 sectors, can't use all of them
Found valid GPT with protective MBR; using GPT

Disk /dev/sdc: 4294967295 sectors, 2047G
Logical sector size: 512
Disk identifier (GUID): 8b974a2e-8823-42a2-ba93-113c35cb58fd
Partition table holds up to 128 entries
First usable sector is 34, last usable sector is 5860533134

Number  Start (sector)    End (sector)  Size       Code  Name

sdc: unknown partition table
 sdc: unknown partition table
 sdc: unknown partition table
 sdc:

I have had the unit for about 2 years and have performed expansion from 1TB to 2TB and later 3TB for each of the 4 drives. Never had any issues and I also replaced 2 failed disks during this time and that also worked fine.

Any clues on what might be the problem?

Is there a way to force the unit to mark the disk as not present/missing?
Thanks.

Hw & Hw Compatibility

Other

14 Replies

Replies have been turned off for this discussion

mdgm-ntgr
NETGEAR Employee Retired
Nov 12, 2012
You should open a support case (see Online Submission link in my sig) and post your case number.
rsilva1
Aspirant
Nov 12, 2012
Thanks, I have just opened a case (19866764). Let's wait and see.
PapaBear1
Apprentice
Nov 13, 2012
When it said disk failed, did it give you any message either on the front screen or in the logs? On some instances the disk may actually be fine, but it may be one sector smaller than the previous smallest disk and will therefore be rejected as being too small. I have hit this several times on some of my 1TB Seagates and on a 500GB Samsung previously.

The only cure I know for it is to back up the entire volume, insert the disk and then perform a factory default. The only preventative I know of is that when you establish the volume and before you add data, is to swap out a known good disk with the one you are going to keep as a spare. Then if you get this message, you can simply factory default and use the good disk that had been removed as a spare.
rsilva1
Aspirant
Nov 13, 2012
I must admit I did not check the logs. Since I have a spare disk, I inserted that one and connected the "failed" one to my PC for diagnostics.
I have just noticed that /var/log/messages on the NAS has not been changed since 30 April, despite the unit being on 24/7.
At the moment, syslog only goes back 2 days and the failure was on the 10th... logrotate does not seem to be working (despite being in cron.daily). There are a syslog.1 and syslog.2 files, but they are zero-length and a few months old...

As for the new disk being smaller, hdparm reports it to have the exact same number of sectors as the other three, so I don't think that is the issue.

/var/log# hdparm -I /dev/sda | grep -i sect
sectors/track 63 63
CHS current addressable sectors: 16514064
LBA user addressable sectors: 268435455
LBA48 user addressable sectors: 5860533168
Logical Sector size: 512 bytes
Physical Sector size: 512 bytes
R/W multiple sector transfer: Max = 16 Current = 0
/var/log# hdparm -I /dev/sdc | grep -i sect
sectors/track 63 63
CHS current addressable sectors: 16514064
LBA user addressable sectors: 268435455
LBA48 user addressable sectors: 5860533168
Logical Sector size: 512 bytes
Physical Sector size: 512 bytes
R/W multiple sector transfer: Max = 16 Current = 0

When I explicitly inserted a smaller disk (to make sure the system was able to recognise the new disk), I got a message saying it was too small.

Netgear support recommended reinstalling the OS, so I will try that before performing a factory default.