Upgrading Disks and Reshaping Time

My ReadyNAS has 2x1TB and 2x4TB drives in it.

I just purchased 2 more 4TB drives to replace the 1TB drives as I need more capacity.

I've swapped in a 4TB drive to replace the first 1TB drive and it took 24 hours to recover the data (orange to green status) but LED display now says "reshaping" (or "syncing" if you looking in admin portal) of the data is going to take another 32 hours. I'm presuming "reshaping" here means spreading the data more efficiently across the 4 drives?

Thing is though that I'm not done yet as still the other 1TB drive to replace.

QUESTION then; can I halt/ignore the reshaping and now swap in the other 4TB drive before reshaping has finished as seems more productive to me to have my full compliment of new drives installed before it does it's final reshaping rather than do this lengthy operation twice.

Can you advise please Netgear Gurus!

Thanks

Julian

Installation & Upgrade

StephenB
Oct 13, 2016
juwi_uk wrote:

So what is your solution?

Only one I can see is to have 2 duplicate NAS's

That is what I do. My main NAS is a pro-6 with a 15 TB RAID-5 volume. I have two backup NAS - an RN102 and an RN202, both with 6 TB + 8 TB jbod (two volumes on each). I learned pre-nas that it was wise to have 3 copies of everything I care about.

I also use crashplan cloud backup for disaster recovery, but I am not so confident in any cloud backup to rely on it completely. At some point I might drop one of the backup NAS and rely on crashplan (or amazon cloud) for the second copy.

I think another option is to use 8 TB usb drives (though you'd need two of course). If a lot of the data is archival, you could use a Seagate SMR drive for that.

juwi_uk wrote:

And I do backup my critcial data, just not all 16TB.

If you have the ability/discipline to separate criticial and non-critical data, then that's a great way to reduce the impact.

I find its difficult to keep "critical" separate from "non-critical" myself. So I just back up everything.

11 Replies

Replies have been turned off for this discussion

mdgm-ntgr
NETGEAR Employee Retired
Oct 12, 2016
First it had to add the disk into the 4x1TB array, now it is adding it to the 2x3TB disk array to create a 3x3TB array so the volume can then expand.

You must wait for the resync to complete and the volume status to return to a redundant state before replacing the next disk. Please be patient.

If you try replacing another disk before the reshape completes your volume would go offline. Don't try this.
- juwi_uk
  Tutor
  Oct 12, 2016
  Not sure where the 4x1TB, 2x3TB and 3x3TB references are coming from as the 4 slots only contain 1TB or 4TB drives and no sign of any 3TB ones!! :0)
  
  Original starting configuration:
  
  Slot 1: 1TB
  Slot 2: 1TB
  Slot 3: 4TB
  Slot 4: 4TB
  
  At the moment it is sync'ing with one disk inserted
  
  Slot 1: 4TB
  Slot 2: 1TB
  Slot 3: 4TB
  Slot 4: 4TB
  
  I will end up with this when last disk is inserted.
  
  Slot 1: 4TB
  Slot 2: 4TB
  Slot 3: 4TB
  Slot 4: 4TB
  - mdgm-ntgr
    NETGEAR Employee Retired
    Oct 12, 2016
    The first RAID layer is 4x1TB as the capacity of the smallest disk you started out with is 1TB.
    
    It uses the remaining 3TB of space on each 4TB disk for the second RAID layer. Expansion will take place when redundant space can be added.
juwi_uk
Tutor
Oct 12, 2016
Is there a way I can end up with just a single layer? Does having the 2 layers become a performance hit over what i would have had if i started with 4x4TB disks?
- StephenB
  Guru - Experienced User
  Oct 13, 2016
  juwi_uk wrote:
  Is there a way I can end up with just a single layer?
  
  Only if you do a factory reset with all disks in place (or destroy the volume and recreate it). Both methods destroy your data.
  
  juwi_uk wrote:
  Does having the 2 layers become a performance hit over what i would have had if i started with 4x4TB disks?
  
  I've never tried to measure it, but I don't think so.
  - mdgm-ntgr
    NETGEAR Employee Retired
    Oct 13, 2016
    I think there would be a small performance hit from having multiple layers (more layers for the system to manage would take a little more effort) but the network connection would be more of a bottleneck most likely so you probably wouldn't notice any difference.
    
    It can make data recovery (we sell contracts for data recovery attempts) a little more complicated (in the unlikely event a data recovery attempt is needed) if you have multiple RAID layers but we do have ways of trying to recover information about the partitions (in the unlikely event we need to do that as that shouldn't be lost). The logs zip download can be useful in such cases too if you happened to have downloaded that.
    
    It would be best practice (though not essential) to download the logs and to download a fresh set after each disk replacement. It would be best to keep the logs zip download in a safe place other than the NAS.
    
    Also when updating the firmware it would be good to follow this: ReadyNAS OS 6: Firmware Upgrade Guide and Tips