NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
Best / quickest / least painful way to “vertically” expand a NAS when having a full data backup
Hello Forum This is – somewhat – a follow up to “vertical expansion of the wrong NAS” (https://community.netgear.com/t5/New-ReadyNAS-Users-General/vertical-expansion-of-the-wrong-nas/m-p/2157308#M47...
Thinking about this a bit more... wanting to use the younger 4 TB drive in addition to the two 12 TB drives might change the calculus some.
One option is to power down the NAS, and replace the three disks - setting the ones you remove aside. Leave the one disk you want to continue to use in the system. Then do a factory default from the boot menu, set up the NAS again, and restore the data from the backup. Personally I'd also upgrade the firmware to 6.10.9.
Advantages are
- gives you a completely clean system
- gives you a fallback if something goes horribly wrong (as the NAS will boot up with the original volume if you power down and reinsert the three disks you remove - leaving the fourth slot empty)
- rebuilds with minimum disk i/o
Personally I don't think (3) is a major consideration in your case, as you are replacing healthy disks and have a full backup.
Hello Stephen, TY for coming back.
First of all an apology, disks were tested using a SATA2 socket – not a SATA3 (I always meant to say that, I just got the digit wrong).
“Any particular reason you are still on 6.10.3 (2020 firmware)?”
“if it ain’t broke, don’t fix it”, my NASs are all working fine. But I take the advice to upgrade to 6.10.9 – I presume that the best time to do it is after I do the “Factory Default”, in the procedure you recommend in your 2nd message.
Also.. change of program, I just bought 2x 12 TB Ironwolf Pro – I should have them Tuesday.
So I wonder if it wouldn’t make more sense to have the 4x 12TB array in the RN424, which is a better machine.
I know that the RN424 is Intel based and the RN214 is arm based, but I do not know if I can simply and safely move an array from an Intel based platform to an arm based platform.
If that is possible, I could:
- Save the config, remove the array from the RN424 (4x 8TB) and put it in the RN214a, and check if it reads the array.
- Refit the RN424 with the 12TB drives, do a factory default, upgrade to 6.10.9 if so recommended (it has 6.10.4 Hotfix 1).
- Copy over all the data from the RN214 which has the original array.
- Then on the RN214a, I would save the configuration and delete the existing volume;
- Maybe do a factory default? Would that be necessary/advisable? (That array always had 8TB drives), then upgrade the FW to 6.10.9;
- Copy over the existing backup of the data which is on my PC (currently testing both 12TB IronWolf, (Victoria v5.37, W-R-V test since yesterday, ~11h to finish)
That.... assuming that an arm platform has no issues with an array created on an Intel platform. If that is NOT the case, I guess that I would follow the same in procedure, but the RN214 ( save config, remove drives, refit drives, factory default, firmware update, copy over the backup.
Same size drives, no resynch necessary, easy.
I simply need to wait for the new drives and test them. Fingers crossed...
Yes, you can move an array form an OS6 ARM unit to an OS6 Intel unit (and vice-versa). There is a very slight difference in the way the drives are formatted for the ARM version, the OS partition is EXT instead of BTRFS, and that will remain, but it's not really a factor you need to worry about.
berillio wrote:
First of all an apology, disks were tested using a SATA2 socket – not a SATA3 (I always meant to say that, I just got the digit wrong).
FWIW, SATA3 has a benefit with SSDs, but SATA2 is still faster than mechanical disks, so that makes no real difference.
berillio wrote:
I know that the RN424 is Intel based and the RN214 is arm based, but I do not know if I can simply and safely move an array from an Intel based platform to an arm based platform.
As Sandshark says, you can. But you do need to uninstall your apps first (and reinstall them after the NAS boots up on the new platform).
Hiya Forum,
Further update. The ebay seller was actually local, so the morning after my post I could collect the disks from his address, came back home and started testing them both, simultaneously. One failed the Quick Test and continued failing the WRV (WriteReadVerify, I presume) test repeatedly; the other one seemed ok, but eventually proved to be only slightly faster than the standard 12GB Ironwolf (non Pro) previously tested, and also has SIX times as many blocks in the 100-250ms range than the non Pro.
I considered that not safe enough and I returned both disks today (the WRV test finished @7:01:05 this Monday morning).
So, back to the original plan, which was the 2x12TB + 2x4TB array,
I will follow the procedure suggested by StephenB the 3rd post.
Update again – plus some kind of snag 😞
I created the 2x12TB + 2x4TB array as planned, went through the single double-resyinch (2x ~36h), upgraded FW and everything worked out perfectly fine.
After copying all the data which was on the NAS original array, I copy-moved there all the data which was scattered around on various 4TB USB drives, other disks on another PC, and the RN424.
Then I started consolidating the dataset moving all various folders in a more organised manner.
This is when I realised that - guess what, I was running out of space on the new array.
I initially panicked, thinking that I had misconfigured the lot and created like 5TB of snapshots (like it happened before, some while ago); (only at that moment I realised - it hadn’t occurred to me before - that the snapshots are created for an entire share and not a subfolder).
So, thinking that the share where I moved the data had enabled snapshots, I created another share, made sure that snapshots were not going to be created there and started moving data in this new share (using the web interface, which I am not familiar with – infact it was probably the FIRST files/folder movement I ever done using it).
Few minutes later I realised that ACTUALLY, I had only 32 GB of snapshots in total in the entire array, and they were in some other shares, so that was NOT why I was running out of space
So I reasoned that, obviously, I did not do my sums right in the first place and underestimated the space I needed (well, I was never running out of space, but I would have ended up with three NASses which would have been 75/80% full, and I would have never had the space to start having some proper backups etc.).
So I decided to get the other 2x 12TB and create the full, proper array. I stopped all that moving to the new share operation, as it was unnecessary at this moment, and started moving the data OFF the new 214a array to the old locations, so I could scrap the new array entirely, create a new one directly with 4x12TB and save all that double resyinching exercise (2x2 resynchs of 36h would have been lengthly...).
This morning I got the 2x 12TB SG IronWolf, from the same “official” (3y warranty) retailer I bought the first couple, planning to ReadWriteVerify test them as soon as possible, and started looking of where to move the data, according to where I had the space, wishing to avoid the USB disks etc.
It was only then that I realised that – SHOCK HORROR – I have lost data, and a conspicuous amount too, ~2.5TB (in one dataset, three years of data, 2018, 2019, 2020 have disappeared entirely and 700GB from the 2021 folder are missing).
And of course I don’t have any snaphots, that folder was (properly, lol) configured NOT to have snapshots.
Of course, if the data was on a normal disk, there are plenty of programs which would allow me to undelete/recover the data. But on a RAID… well, I know that there a few programs which can reconstruct failed/corrupt array (never been able to recover any data with the shareware versions, tho’), but this is a different situation, the array is (presumably) not corrupt, I simply cancelled/aborted the “moving” operation without realising that ONLY the data which had been transferred would have survived (if this is what happened). This is why one should have a snapshots/backup policy in place.
At moment I am moving data out of the array. That may take over a day or two if not more. I am also testing the new disks, circa 30h remaining.
Once all the (transferred) data has been moved back, since I still have the original full backup of the data which was on the 214a before doing anything, I could do a factory default start from scratch again. Unless there is a way to recover those lost subfolders, that is.
Incidentally, all the data which has been lost came out of the RN424, and I am not moving any data there yet (but I might have written ~100GB of new data in the last few days). Needless to say, there are no snapshots, but 5.1TB of free space, ~25%. If I wish to leave that space alone, I have the 3x 4TB disks which were de-commissioned from the 214a which could be wiped clean if necessary; I also have 3.3TB free-able on a PC and ~2.95TB on an USB.
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
