NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
vertical expansion - of the wrong nas?
Hello Forum, I have an issue, but describing became War and Peace (600 pages of it), so this is the short of it I have three NASs, (say A, B, C) (4x 4TB, 4x 4TB, 4x 8TB). A & B run OS-6.10.3, C ...
berillio wrote:
A 4Tb disk failed on NAS A. I can replace it with a new 8TB I have – and start expanding, which is good because I need more storage space.
but
That is NOT the NAS which needs expanding. It is NAS B which needs expanding; it has one aging disk but also three newish one (april 2020).
What can/should I do?
I tend to favor minimal steps - so I'd normally just replace the failed disk in A with the 8 TB drive. Since it isn't going to expand, you can replace the 8 TB drive with a 4 TB later on if you want to.
However, your plan A would also work. Generally I test my drives in a PC before inserting them into - first running the full non-destructive generic test, and following that up with the full erase/write zeros test. I'd recommend doing that on the 4 TB drive you want to re-use. Erasing it will also avoid any confusion in NAS A when you add it.
It would be good to make sure you have a backup of the unique files of each NAS before you manipulate its drives. So if you want to expand anyway, getting a second 8 TB drive for that purpose makes sense.
I'm not sure what your long term plan here is. I suggest designating one NAS as the "primary" NAS, and putting all the content you have on that NAS. Then use the other 2 as backups. You can over time expand them all to have the same size (giving you two full backups), but in the beginning you can back up some shares on each (giving you one full backup between the two NAS).
If you go with that suggestion, then you first should figure out what capacity you want in each of the NAS.
berillio wrote: I can expand NAS A. Then move the ENTIRE contents from NAS A to NAS B and viceversa. Lengthy (~8TB each way, incredibly messy, and (if I understood right, as they are 90% a mistery to me) all the snapshots will get jumbled up and possibly useless.
You might also want to re-think how much retention you really need in the snapshots. For me, snapshots are a way to recover from user error. If retention is too short, then you might not realize that you need to recover something until it's too late. If the retention is too long, then you end up with a lot of disk space, and a lot of fragmentation in the main shares. I tend to use 3-month retention on the snapshots (though some shares have shorter retention).
Just to clarify this. If you
- use NAS backup jobs to copy everything on A to B
- then use NAS backup jobs to copy everything on B to A
A and B will have identical content in the main shares, but A and B will retain their original snapshots. The original B snapshots won't be on A (and vice versa).
FWIW, I do use NAS->NAS backup. My RN526x is the primary NAS. I do share-by-share backup to the other NAS (running daily), with daily snapshots enabled on each NAS.
The snapshots on the backup NAS are similar to the primary, but not identical. If a folder is renamed on the main NAS, then the rsync backup ends up doing a copy/delete. So the snapshots will reflect that (using more storage on the backup). But it is close enough for my purpose. If I rename a really large folder, I can always go into the backup NAS and rename the folder there as well.
Plan 1 has the risk that the older drive will fail during the NAS2 re-sync. Just how old and whether there are any SMART errors are factors in the likelihood of failure. How long can you afford to have NAS1 non-redundant (which you'd base on how much "churn" there is on it, how well backed up, and whether or not you have any SMART errors on the remaining drives. If you can wait for the time it takes to expand NAS2, then replace the older drive in NAS2 with 8TB. Then, when that sync finishes, replace a newer one and move the newer 4TB into NAS1. Before moving it, you may want to zero it on a PC, especially if the volume name on NAS2 is the same as on NAS1.
thanks Stephen B and Sandshark.
My fault, it was in the "war & peace" but not in the summary.
NAS Ais OFFLINE
That is the NAS with the failed drive - actually it has not failed yet, I started receiving mails 32-40-48 errors, I saved the logs and switched it off before it did.
NAS A is rarely accessed and written to, it can wait offline a week or two.
Actually, if I power it on to copy the data BEFORE doing any volume rebuilding, would it be better to do it with the failing disk removed (it is already out), or power it back on, and hot remove it?
I had a disk failure initiating a daisy chain failure fo an healthy disk, I would prefer REMOVE the disk before it the NAS failed it
berillio wrote:
Actually, if I power it on to copy the data BEFORE doing any volume rebuilding, would it be better to do it with the failing disk removed (it is already out), or power it back on, and hot remove it?
I had a disk failure initiating a daisy chain failure fo an healthy disk, I would prefer REMOVE the disk before it the NAS failed it
The puzzle here is whether you have multiple disks on the edge or not.
If only a single disk is at risk, then removing it is fine. But if multiple disks are stuggling, then I think it's best to leave them all in place. The theory there is that the bad sectors aren't likely to overlap, so all the data is still recoverable.
Personally I've never seen a case where a failing disk provoked a failure on a second drive.
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
