My system is protected by a UPS and all data is backed up elsewhere. This NAS keeps a daily mirror or another NAS and is only really used to recover from "mistakes"
I have a unit with 4x3TB disks and 1x4TB disk giving me 10TB raid 5 I would like to replace the disks with now cheaper larger disks. I propose to insert a 16TB in slot 5 as a stand-alone volume and copy the date from the raid in volumes 1-4 When I have a copy. I plan to delete the existing data on disks 1-5
Insert say 3x16TB and create a raid 5 copy the data back and expand the raid to 4x16TB or maybe 2 redundant disks.
DOES THIS SOUND SENSIBLE / SANE?
Or should I do something else
I don't know yet if it will accept 16TB disks or what is the largest. I just happen to have quite a few 16TB drives waiting to be repurposed.
I don't know yet if it will accept 16TB disks or what is the largest. I just happen to have quite a few 16TB drives
There is no known limit to the max disk size. You do want to use enterprise class drives, or NAS-purposed drives that don't use SMR. If these are shucked drives or drives identified as "Archive" drives, you should first make sure they are not SMR.
I propose to insert a 16TB in slot 5 as a stand-alone volume and copy the date from the raid in volumes 1-4 When I have a copy. I plan to delete the existing data on disks 1-5
Insert say 3x16TB and create a raid 5 copy the data back and expand the raid to 4x16TB or maybe 2 redundant disks.
If you do this, you need to switch to FlexRAID first. Then you insert the 16 TB drive into slot 6 (NOT slot 5), and create a JBOD volume on it. Then create shares and copy the files. You'll need to use temporary share names, as they need to be unique to the NAS. After the copy is complete, you can just destroy the existing data volume and remove the disks. Note you can power down the NAS after you destroy the data volume, and then move the 16 TB drive to another slot.
Once that is finished, you can add a second 16 TB drive for redundancy (changing to RAID-1) and additional 16 TB drives later on for extra space (changing to RAID-5).
The other option is to just stick with X-RAID and hot-swap the 16 TB drive (removing one of the 3 TB drives with the NAS running, and hot-inserting the 16 TB drive in its place). The RAID array will resync and expand by 1 TB (due to the 4 TB drive in slot 5) - giving you 13 TB (11.8 TiB). When the resync is complete, hot-swap one of remaining the 3 TB drives with a 16 TB model. You'll end up with 2x3TB+1x4TB+2x16TB X-RAID, with a total volume size of 26 TB (23.6 TiB).
This alternative is less work. Whether it is "better" depends on your reasons for making this change (in particular whether you are concerned about the condition of the disks you are currently using).
Note X-RAID isn't proprietary. It is just management software built on standard Linux RAID that simplifies expansion.
Thanks - I'm still trying to decide should I buy another in support Synology (or other make) chassis and retire the Netgear. - Maybe put a 3rd party Open source software on the Netgear if that's even possible, Is there some sort of re-sync/rebuild calculator for Netgear?
Recently I added a new disk to a Raid 1 set on TrueNas 6GBs 7200rpm and it took 30 hours to rebuild the disk for 16TB that was shorter than it's estimate of 2.5 hrs per TB.
My main setup (I'm a retired IT guy) 2 Synology 20TB nas 1 wakes up on a Sunday and mirrors to primary then backups to a tape robot each month
2 Netgear NAS 10TB having a subset of files from the Synology - mostly TV shows
Various other random NAS that I don't really care about that hold current TV shows
A Truenas system in an old Xeon server that I use to test new disks before I put them in a NAS that I'm going to rely on.
I'd like to De-dupe some of this but don't want to break the 3-2-1 rule.
When you create a new volume , the old one will be what I call the "primary volume" (I've never seen an "official" name for it), still containing all apps and Home Folders. When you destroy that primary volume, the other has to become primary. But, sometimes that seems to go wrong, and there are a couple things that never happen.
In addition, your new volume will have to have a different name, and it's a VERY difficult task via SSH to change that. So if you have mapped drives or share call-outs that include the volume name on the NAS itself or other devices, you'll have to update those.
If you have and use user Home Folders, those will not be transferred. The shares themselves won't even be created until the next time the user logs in, and the files will never be transferred. So, you'll have to deal with those yourself. It's a bit of a Catch-22 since you cannot create home folders on the new volume while the old is still mounted in order to transfer the shares. There are some tricks you can use, so let us know if this affects you (I suspect it does not, so won't go into details unless you ask). One of them you can do as a precaution is to EXPORT instead of destroy the old volume. The new one still becomes the primary, but you can re-install the old one (with power off) and it will then be mounted as secondary on power-up. You cannot re-mount a destroyed volume.
Additionally, apps will likely not be properly transferred, though it appears they are supposed to be and maybe that's been fixed in an OS update since I last experimented with this stuff. See How-to-save-your-apps-when-destroying-your-main-volume-OS6 on how to handle that.
As for the chance it simply won't go well, the fact that you have a backup NAS is good. Make certain it's up to date. Should that occur, you have a recovery method. Since this happened to me twice while I was doing a lot of experimenting with just this sort of thing, I don't know if that was because that was a typical failure rate of the process (2 out of 20ish) or was due to the numerous experiments I was performing confusing the OS.
You may want to consider if it's not actually better in the long run to simply shelve the old drives and start fresh with the new, since you have a backup NAS from which you can copy everything. If you find you've forgotten something, you can just swap out the drives and the old volume will be back just as everything is now.
IMHO, you 316 is still a very viable backup to your Synology. For a home user who has not opened any access to the outside world, I really think any security risks of the outdated OS are minimal.
Is there some sort of re-sync/rebuild calculator for Netgear?
I'm not sure if it's still available.
But the capacity rule for single-redundancy XRAID is pretty simple - "sum the disks and subtract the largest". That will give you capacity in TB (the NAS UI tells you capacity in TiB).
As far as the RN316 goes, as I said above the easiest path is to hot-swap two of the 3 TB drives with 16 TB drives - one at a time of course , and waiting for the NAS to synd the first before hot-swapping the next.
Netgear never provided a sync time estimator, but in general the NAS does need to access every sector of in the expanded volume to sync. If you just expand the current volume, The first sync requires 17 TB of disk I/O. The second will require 42 TB, so quite a bit more.
The first requires less I/O because the NAS can't use 12 TB of space on the 16 TB drive until a second one is added. So 3*5 TB of I/O to resync the initial volume, plus another 2 TB to expand that volume by 1 TB (using the currently unused space on the 4 TB drive). The second requires accessing (either reading or writing) 2x3TB+1x4TB+2x16TB, so 42 TB all together.
