× NETGEAR will be terminating ReadyCLOUD service by July 1st, 2023. For more details click here.
Orbi WiFi 7 RBE973
Reply

Re: ReadyNasUltra 4 dead

garethjjones
Aspirant

ReadyNasUltra 4 dead #26795494

One of the disk was showing increasing Smart errors, so I ordered a replacement disk (ST3000VN000) -to replace the failing disk. No big deal I thought.

 

I had assumed that the failing disk was disk4 - a 3Tb Seagate disk I'd ad some smart errors on before - as I didn't read the emails closely enough.

 

When I was getting an email every few minutes, I shut down the NAS to wait for the replacement disk. It arrived, and I replaced disk 4 with the new disk, and booted up. The system stayed at booting for some time - whilst I now read the message which made it clear that the failing disk was Disk 1. I powered down the NAS.

 

I now replaced the original disk 4 - a ST33000650NS and rebooted. This appeared to work, and the Nas came up and the SMART error messages flooded in. 

 

I decided to hotswop the failing disk, Disk 1 (ST2000VN000-1H3164). Then it all started to go wrong.....

 

I get the messages in the following order:

Disk removal detected - Disk1

Raid event detected - Data volume will be rebuilt with disk 4

New disk detected Disk 1

Raid event detected - Data volume will be rebuilt with disk 1

 

Then nothing happened - no raid sync on the volume, and most of my files dissapeared. I downloaded the logs using frontview, and then looked at the volume status - which showed that the X-raid was dead, and that I had two spare disks. That was midnight, so I left the NAS until this morning. (No sleep - a lot isn't backed up.....)

 

This morning the NAS appeared to have dies - though I could look at front view, I couldn't refresh it - and the NAS front panel couldn't be brought to life. I've now powered it down and asking for help. What's the next step?

 

I wonder if I can replace the faiing Disk1 back in order to rebuild the raid array,  and when that's done replace it, taking any data hit as getting the Raid array back is the priority.

 

 

 

 

 

Model: ReadyNASRNDU4000|ReadyNAS Ultra 4 Chassis only
Message 1 of 13

Accepted Solutions
garethjjones
Aspirant

Re: ReadyNasUltra 4 dead #26795494

Netgear support managed to rebuild the array remotely, using mdadm and then mount the array.

Powered off, removed the bad disk 1 and rebooted with only disk 2,3 and 4 and then could access with fontview.

Put the new disk in disk1 slot, and the raid array has re-synched and we're now back online, all data intact (AFAIK).

 

Now looking at implementing crashplan on the Ultra4 - though not sure how long it would take to upload data!

 

 

View solution in original post

Message 12 of 13

All Replies
StephenB
Guru

Re: ReadyNasUltra 4 dead

If you have a backup (or are ok with losing data) then I'd suggest doing a factory reset, which will do a fresh factory install.

 

You'll need to reinstall any add-ons, and reconfigure the NAS, and restore the NAS data.  This will take a while, but will get everything back to a known/clean slate.

 

 

Another option is to contract with Netgear.  With per-incident support they might be able to remount the volume.  Data Recovery would be its own (and more expensive) contract.

Message 2 of 13
garethjjones
Aspirant

Re: ReadyNasUltra 4 dead

Not all the data is backed up - that's the issue - so I'm not really happy with losing the data.

 

I've created a support case 26795494, and uploaded the logs - though if anyone else has any ideas I'm happy to hear them. 

 

It's pretty difficult to find out how much this service costs - only by searching the forums, and then it's "about"

 

 

Message 3 of 13
JennC
NETGEAR Employee Retired

Re: ReadyNasUltra 4 dead

Hello garethjjones,

 

The per incident contract is GBP60 good for 1hr but this will possibly fall under data recovery service. Let's see what support center can offer.

 

Regards, 

Message 4 of 13
garethjjones
Aspirant

Re: ReadyNasUltra 4 dead

Thanks - that's not to bad...doesn't seem like data recovery to me 😉

 

Message 5 of 13
StephenB
Guru

Re: ReadyNasUltra 4 dead


@garethjjones wrote:

Thanks - that's not to bad...doesn't seem like data recovery to me 😉

 


Per-incident support does not include data recovery.   However, they might be able to successfully mount your raid array, and they might be able to give you some assessment on your prospects of a successful data recovery.

Message 6 of 13
mdgm-ntgr
NETGEAR Employee Retired

Re: ReadyNasUltra 4 dead #26795494

Looking at the logs it looks like a data recovery situation to me. If your data is important then you need to backup your data. In rare cases or e.g. if alerts about disk errors are ignored for too long you can get data loss.

One disk had a lot of command timeouts and another a lot of uncorrectable errors. Two disks have dropped out of the array. 

The NAS seems to think that there are five disks in some arrays (though they are still considered 4 disk arrays). In one of the layers one disk is marked as failed and two as spare (though it is a 4 disk layer).

If you purchase per incident support contract and that does not resolve it you may need to purchase a data recovery contract in addition to what you will have already purchased. The support may be unsuccessful.

Message 7 of 13
sportsnapper
Aspirant

Re: ReadyNasUltra 4 dead #26795494

Thanks mdgm, 

 

Disk 4 had command timeouts for a while - though not very often - which is why I replaced that with the new disk first rather than the disk1, which was the one that with uncorrectable errors.

 

I've found the majority of the really important work data on a backup disk - though (and I guess like many people) I'm not entirely sure exactly what is on the NAS. I'm going with the per incident suport in the hope that we can get the array back, and then I can take a decision if I want to try to recover data further. There's a lot of films and tv shows that I've never watched, and prob never will - not becasue they're lost but we don't have enough time.

 

Which logs did you look at to work out the info about the array - the expansion log?

 

 

Message 8 of 13
sportsnapper
Aspirant

Re: ReadyNasUltra 4 dead #26795494

Arggh - i've ended up with two id's on the forum..... 😞 

Message 9 of 13
mdgm-ntgr
NETGEAR Employee Retired

Re: ReadyNasUltra 4 dead #26795494

I looked at the status log to see the warnings that were given and looked at mdstat.log to see a quick overview of the current state of the array.

Message 10 of 13
garethjjones
Aspirant

Re: ReadyNasUltra 4 dead #26795494

Thanks - useful for the future hopefully.

 

I've updated my case, and now waiting for support to get intouch about rebooting into tech support mode. Not sure which disks should be in the unit at that point - though I think they may be on a different timezone so wondering when they'll get back to me. 

 

 

Message 11 of 13
garethjjones
Aspirant

Re: ReadyNasUltra 4 dead #26795494

Netgear support managed to rebuild the array remotely, using mdadm and then mount the array.

Powered off, removed the bad disk 1 and rebooted with only disk 2,3 and 4 and then could access with fontview.

Put the new disk in disk1 slot, and the raid array has re-synched and we're now back online, all data intact (AFAIK).

 

Now looking at implementing crashplan on the Ultra4 - though not sure how long it would take to upload data!

 

 

Message 12 of 13
StephenB
Guru

Re: ReadyNasUltra 4 dead #26795494


@garethjjones wrote:

 

 

Now looking at implementing crashplan on the Ultra4 - though not sure how long it would take to upload data!

  


Of course it depends on your uplink and the congestion on the path to their servers.  My backup speeds vary quite a bit - usually between 16-75 mbits.  Du-duplication helps by reducing the data transfer - especially on my image backups.

 

If I were doing it over today, I'd install crashplan on a PC and mount the NAS as a network drive letter.  

 

One reason is that over the past year they have upgraded their platform requirements a few times, and I am guessing that will continue.  There were some bumps in the road over the summer (where new Java and a new glibc were needed on the NAS).  

 

Another reason is that crashplan deduplication can use a lot of memory.  If crashplan doesn't have enough memory, it basically stops working.  I believe an ultra-4 has only one memory slot, and is limited to 2 GB ram max.  My pro-6 can (and does) use 8 GB, but the java vm is limited to about 3.7 GB.  

 

FWIW, there is a way to turn off de-duplication if you hand-edit the config file.  But Crashplan really doesn't want you to do that (one reason is that it increases their cloud storage costs), so my guess is that they will close that option off.

 

 

 

Message 13 of 13
Top Contributors
Discussion stats
  • 12 replies
  • 3516 views
  • 0 kudos
  • 5 in conversation
Announcements