NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
ReadyNAS 314 - a year of hell and now lost 10years of data!
I bought a ReadyNAS 314 when they were new in the market back in April 2013. I bought it partly because I wanted to safeguard my increasing data storage with redundancy in case of disk failure, and also because of the fancy Surveillance add-on that apparently came pre-installed on it (according to the Netgear ReadyNas website)
I bought a diskless 314, 4x Seagate ST3000DM001 HDDs (on the hardware compatibility list), and a nice Panasonic outdoor PTZ wireless camera (on the hardware compatibility list for Surveillance)
Total cost = LOTS!!!
Turns out the Surveillance add-on wasn't actually written for the new ReadyNas boxes yet, and I spent 6 months being a guinea-pig for the tech support, going from one non-working version to another, upgrading with Beta software, uninstalling it, rebooting ReadyNas, restoring the software on it, and eventually an almost working version was put on my box, just in time for me to go on holiday and come back to find we'd been broken into! Nothing recorded on the ReadyNas as the Surveillance add-on had stopped working again!!!
Anyway, through the last 6 months I have had nothing but disk errors reported by the ReadyNas by e-mail. This has been "being looked at" by support for a few months but I was told not to worry about them and they'd figure out what was reporting them or causing the errors. Either way I wasn't going to have to replace all of my drives 10 times a day like it suggested!
So I didn't!
I went away in March for a month, and whilst away amongst the bombardment of e-mails about disk errors, I received one saying "Volume is DEGRADED", nothing before that to suggest 1 disk had gone down, then another. Just that e-mail!
Stupidly I thought, oh that'll be a disk failure then and the box is going to shut down/protect itself and I'll replace the bad disk which is under warranty when I get home.
I come home to find flashing lights on the ReadyNas, I cannot log onto the ReadyNas console to see what is going on, so I contact support.
They log on to it, cannot get any logs off it as they all appear lost (with the data), and tell me that 2 of my disks died!!
Looking back over my e-mails, both disks must have died in less than a 45minutes timespan!!
Really, am I that unlucky? Can 50% of my less than a year old HDDs really die that close to each other?
Netgear of course are not going to pay to recover the data if it is even recoverable. Seagate will replace the disks under warranty but then the data is definitely gone, and a 3rd party want best part of £1000 to recover it for me!
So, shall I just throw the whole lot in the bin and NEVER buy Netgear products again? (Currently have 4x ReadyNAS rackmounts and 2x ReadyNas Pros at work, to be replaced soon as cannot risk something like this happening there!)
Shall I spend more money on an awful product and get the data recovered, replace the disks, and then maybe watch the same happen again?
Or should I jump up and down until Netgear accept that as a customer I have had a really, really bad year dealing with their unfinished/untested software, inadequacies in support, and the fact that it would appear my data was safer for the last 9 years NOT using Netgear products than it was spending over £1k on their really safe products!)
I would love to hear a response from Netgear...
I bought a diskless 314, 4x Seagate ST3000DM001 HDDs (on the hardware compatibility list), and a nice Panasonic outdoor PTZ wireless camera (on the hardware compatibility list for Surveillance)
Total cost = LOTS!!!
Turns out the Surveillance add-on wasn't actually written for the new ReadyNas boxes yet, and I spent 6 months being a guinea-pig for the tech support, going from one non-working version to another, upgrading with Beta software, uninstalling it, rebooting ReadyNas, restoring the software on it, and eventually an almost working version was put on my box, just in time for me to go on holiday and come back to find we'd been broken into! Nothing recorded on the ReadyNas as the Surveillance add-on had stopped working again!!!
Anyway, through the last 6 months I have had nothing but disk errors reported by the ReadyNas by e-mail. This has been "being looked at" by support for a few months but I was told not to worry about them and they'd figure out what was reporting them or causing the errors. Either way I wasn't going to have to replace all of my drives 10 times a day like it suggested!
So I didn't!
I went away in March for a month, and whilst away amongst the bombardment of e-mails about disk errors, I received one saying "Volume is DEGRADED", nothing before that to suggest 1 disk had gone down, then another. Just that e-mail!
Stupidly I thought, oh that'll be a disk failure then and the box is going to shut down/protect itself and I'll replace the bad disk which is under warranty when I get home.
I come home to find flashing lights on the ReadyNas, I cannot log onto the ReadyNas console to see what is going on, so I contact support.
They log on to it, cannot get any logs off it as they all appear lost (with the data), and tell me that 2 of my disks died!!
Looking back over my e-mails, both disks must have died in less than a 45minutes timespan!!
Really, am I that unlucky? Can 50% of my less than a year old HDDs really die that close to each other?
Netgear of course are not going to pay to recover the data if it is even recoverable. Seagate will replace the disks under warranty but then the data is definitely gone, and a 3rd party want best part of £1000 to recover it for me!
So, shall I just throw the whole lot in the bin and NEVER buy Netgear products again? (Currently have 4x ReadyNAS rackmounts and 2x ReadyNas Pros at work, to be replaced soon as cannot risk something like this happening there!)
Shall I spend more money on an awful product and get the data recovered, replace the disks, and then maybe watch the same happen again?
Or should I jump up and down until Netgear accept that as a customer I have had a really, really bad year dealing with their unfinished/untested software, inadequacies in support, and the fact that it would appear my data was safer for the last 9 years NOT using Netgear products than it was spending over £1k on their really safe products!)
I would love to hear a response from Netgear...
84 Replies
Replies have been turned off for this discussion
- Yeah, if disks can't be read then in some cases swapping mechanical parts using parts taken from a working drive of the same model may help.
As you can imagine such a labour intensive process is one of the reasons that professional data recovery costs a lot of money. Backing up data can be expensive particularly if you have a lot to backup, but the cost of backups has to be weighed against the cost of professional data recovery and how much one values data.
I'm not sure whether I'd take the controller from one failing disk and put it on another, but then I guess it depends on what in the disks has failed. - Disk 4 seems to be cloning ok - well it has definitely been running for a lot longer than disk 2 managed, and there's only the infrequent clickedy-click coming from it.
Here's the status at the moment:GNU ddrescue 1.16
Press Ctrl-C to interrupt
rescued: 13600 MB, errsize: 513 GB, current rate: 0 B/s
ipos: 442674 MB, errors: 841, average rate: 447 kB/s
opos: 442674 MB, time since last successful read: 11 s
Copying non-tried blocks...
Is this good, bad, gonna take much longer?
What goes up and when? is 513GB of errors going to go downwards as it recovers stuff, do I need to run the other 11 commands that dorp the skip size down and do it in reverse? (that's another 11 runs through!) Does it get quicker on each pass as it only has to try the bad parts? Do I have to many to make it worthwhile? etc...
Here's what is currently still running from this morning:ddrescue -r 1 -a 10485760 -b 4096 -d -f -K 65536 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (sector size=4096, forced, min readrate=10485760, skip size=65536, retries=1)
Here's what is left to run!:ddrescue -R -r 1 -a 10485760 -b 4096 -d -f -K 65536 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (and again in reverse)
ddrescue -r 1 -a 10485760 -b 4096 -d -f -K 16384 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (skip size down to 16384)
ddrescue -R -r 1 -a 10485760 -b 4096 -d -f -K 16384 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (and again in reverse)
ddrescue -r 1 -a 10485760 -b 4096 -d -f -K -A 4096 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (skip size down to 4096, retry all failed/skipped sections)
ddrescue -R -r 1 -a 10485760 -b 4096 -d -f -K -A 4096 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (and again in reverse)
ddrescue -r 1 -b -d -f -K 1024 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (skip size down to 1024, no min readrate)
ddrescue -R -r 1 -b -d -f -K 1024 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (and again in reverse)
ddrescue -r 1 -b -d -f -K 256 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (skip size down to 256)
ddrescue -R -r 1 -b -d -f -K 256 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (and again in reverse)
ddrescue -r 1 -b -d -f -K 64 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (skip size down to 64)
ddrescue -R -r 1 -b -d -f -K 64 /dev/sdb /dev/sdc /mnt/hdd0/btrfsd4log.txt (and again in reverse)
If someone could give me a quick run-down of the process and what I would see in numbers that would be great. - I don't think this looks very good, but this first pass is 10MB reads so it will tend to skip a lot of data when their are read errors. You need to try the other passes to get more data off the disk. Unfortunate that you had two bad disks, but not unheard of. They probably came from the same lot. They were new drives right? Seagate had a contamination problem that caused some early end of life issues a year or two ago.
On a positive note, your problem appears to be the disks and there is a good chance the NAS is fine. It is theoretically possible that a bad controller could do write-longs to corrupt HW crc's and cause the disk to have subsequent read errors, but I've never seen that happen unless we're doing it on purpose to create disks with read errors for test situations :-).
Do have disk smart info for either of the disks you could post? - So, I return home to find that it's pretty much gone the same way as disk 2:
Current status
rescued: 14871 MB, errsize: 2985 GB, current rate: 0 B/s
ipos: 597228 MB, errors: 623971, average rate: 50270 B/s
opos: 597228 MB, time since last successful read: 4.2 h
Trimming failed blocks...
Is this as to be expected?
Should I just leave it for another.... day/week/month/year?
It's been 4.2hrs since the last successful read!
Will that error size drop down (it's 99.5% errors!) dsm1212 wrote: They were new drives right? Seagate had a contamination problem that caused some early end of life issues a year or two ago.
Yes all delivered in a box snuggly fitted together from Computer 2000 I believe (through a 3rd party reseller), they were bought March/April 2013
As for SMART data, I think I can get that when this pass has finished (if it ever does). any particular command best to get it?- On linux you can get smart output with:
smartctl -a /dev/sda (or whatever your device is)
If it hasn't had a successful read in 4.2 hours I think I'd give up. Drive timeouts are typically 30 seconds for a read. I doubt ddrescue is doing that many retries so basically it's skipping 10MB every 30 seconds. You've got about 1/6th of the drive done so if no other reads work, worst case you've got 250000 timeouts ahead of you (125000 minutes yikes!). Since both drives did this, there is a very slim chance that this form of read/write exercise is exposing either some bug in the firmware or exacerbating the hw problem. You may want to power cycle the drive and try different ddrescue settings. Maybe smaller read/writes. It's worth a shot.
steve - Interrupted ddrescue, powercycled drive, pc, drive again, restarted etc. Can't get drive to be seen in Gentoo :-(
- Well that is truly unfortunate, I feel for you :-( If both drives are THAT broken, it's very hard to do anything. I'm not sure a controller swap will accomplish anything here, but at least it can't hurt trying.
- Disks are under warranty (warranty is typically 2 or 3 years for most disks) so it will void support from the manufacturer...
With unreadable disks you have no choice than to ask professionals to tear them appart and read them like CD but that will cost a lot since it must be done in a dust-free environnement with appropriate hardware. - If I get a 3rd party to tear them apart to attempt a recovery, will that void my disk manufacturers warranty?
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
