Inexpensive NAS for over 4TB backups - recommendations?

The RN102 is the Netgear equivalent - same processor speed, and similar performance. I use mine for backup (with 2 WD30EFRX drives installed, jbod). The RN102's frontview backup works well with my Pro-6 as the source. (I'm not suggesting that the Synology won't). The main issue with the RN1xx and RN3xx is that OS6 still has some issues that need fixing.

If you are thinking using RAID-0 to spanning the volume over 2 disks, then I strongly recommend against it. RAID-0 spanning drives is too fragile - any failure destroys the volume. JBOD or RAID-0 that doesn't span volumes are the same thing as far as I am concerned. The main issue with multiple backup volumes is that you need to manually balance the storage occasionally.

Overall I like backing up to another NAS - it keeps the backup device completely separate from the source.

If you need more capacity, then maybe get a four-slot unit - they would be more cost effective.

Of course! Didn't even think of a smaller ReadyNAS. The RN102 is actually cheaper than the Synology so it makes perfect sense.

So if I understand you correctly, best to set up as a multi-volume JBOD, and schedule and run Frontview backup from the RN102 rather than the other way around (scheduling and running from Pro)?

Is there any safe way to create a single large volume on a two-bay system like this, so I can avoid having to manually split data across two volumes?

Yes. One caveat is that the RN102 is using btrfs as the file system. I have had no issues with that, but it is a bit cutting-edge for operational use. It's snapshot feature is quite different from the Pro, and in my opinion more useful. I'd leave the anti-virus scans off (doesn't make much sense on the backup, and slows things down).

Also, OS-6 currently lacks disk spin-down. My guess is that they will add it shortly, but of course that gambling on futures.

dhl wrote:
...So if I understand you correctly, best to set up as a multi-volume JBOD, and schedule and run Frontview backup from the RN102 rather than the other way around (scheduling and running from Pro)?

Yes. I always run the backups on the destination NAS, it seems better especially if you are combining with scheduled power-on. As always, be careful to not to be aggressive on power-down, since the NAS will shut down even if the backup is not complete.

BTW, the RN102 also has WoL, so you could alternatively boot it up from one of the Pros on schedule. I haven't played much with that feature.

dhl wrote:
Is there any safe way to create a single large volume on a two-bay system like this, so I can avoid having to manually split data across two volumes?

RAID-0 is the only way. If you are wanting long-term archival, it is not safe. If you are just wanting a current snapshot of the pro, it might be good enough.

I recommend a UPS in any case, but especially if you go with RAID-0.

StephenB wrote:
RAID-0 is the only way. If you are wanting long-term archival, it is not safe. If you are just wanting a current snapshot of the pro, it might be good enough. I recommend a UPS in any case, but especially if you go with RAID-0.

We keep everything on HCL-compatible UPS (with USB monitoring).

On main goal with the backup system is not archiving. We mainly want data protection and recovery in case our main NAS fails or (as in the current case) we need to do a factory reset for whatever reason.

StephenB wrote:
BTW, the RN102 also has WoL, so you could alternatively boot it up from one of the Pros on schedule. I haven't played much with that feature.

Would be great if the some basic scripting were possible with backup scheduling - i.e. run backup on day/time, then power down when finished. Feature request! :wink:

It seems like we'd get the most flexibility with a RN104 as we could run that in X-RAID2 single-redundancy mode and still have enough future capacity to back-up the Pros in dual-redundancy, even if we fill all the Pro's slots. I'll have to see if we have enough budget to do this.

Re: snapshots - I've never used this feature. I get the impression it works a bit like Time Machine on the Mac, copying differences but making each instance look like a full backup so it takes much less time than a regular back-up?

Do snapshots have any big advantages over regular Frontview incremental backups?

Can snapshots be used for full volume recovery (i.e. recover C volume)?

Can snapshots be saved to a backup drive instead of the source drive?

btrfs snapshots are described here: https://btrfs.wiki.kernel.org/index.php ... napshot.3F

If you do incremental backups with snapshots on, then the previous versions of the folder can still be accessed. This does take space of course, and you may not need them. The main value is that it gives you some ability to recover from user error.

It's a btrfs file system feature, so the snapshots have to be in the same volume as the folder. OS6 creates each share as a btrfs subvolume, so the snapshots are essentially part of the share. One consequence is that moving files between shares in OS6 is a copy/delete operation - not as fast as on the Pro.

When restoring, the snapshot folders would be included (so you might need to delete them before restoring).

Interesting. This this the key quote for our purposes:

Note that a snapshot is not a backup

So probably not what we want.

I'll dig deeper into btrfs and OS6 to see if anything else stands out. I'm sure we can live without the spin-down option. We'll make sure to schedule power on/off with enough headroom to insure finished backups, or use WoL and scheduled power down.

So depending on budget, looks like a RN102 or RN104 should do the trick.

Thanks again for your help!

I agree, the snapshot is not a backup and you shouldn't really think of it that way.

A reasonable way to use them is to (a) set the backup job to delete files that are deleted on the source, and (b) turn snapshots on the destination.

Then it works a bit like the recycle bin. I can still get back files that I accidentally deleted on the Pro (assuming I act reasonably quickly), but the main backup folder matches what's currently on the Pro.

The upgrade on our LA-based Pro system is complete and I've begun backing up to our 104 backup NAS.

I set up two backup jobs:

Job 1)
source - volume C (ReadyNAS Pro)
destination - remote NSF volume /data/Backup (on the ReadyNAS 104)

Job 2)
source - volume C (ReadyNAS Pro)
destination - remote RSYNC volume Backup (on the ReadyNAS 104)

I tested and verified both jobs were set up and working properly

I ran Job 1 as a full backup. It took about three days and then failed with an unknown error. There is no information in Job 1 log on the cause of the error. I checked the backup volume on the 104 and 2.94GB of 3.06GB has been successfully copied. I did a spot verification of the files and they all seem good. All critical shares and directories seem to have been backed up fully.

I don't know why NSF failed but the 104 is still responsive and at less than 50% capacity. I also upgraded to 6.1.5 before starting this process.

Since 90% of the backup was complete, I started the RSYNC job to initiate regular incremental backups.

Questions -
When setting up a backup job, it appears that a full backup must be performed at least once at the start. According to my logs, RSYNC is doing a full back-up but is first sending an incremental file list of every file in volume C.

Is this normal?

Is RSYNC actually doing a full back-up all over again? Or is it simply building a list to compare against what's already on the backup NAS? I'm hoping it copies only the files that didn't get copied with the first NSF backup.

Thanks!

I had a similar issue 2 years ago dhl and this web page helped me to avoid doing another FULL backup:

http://carlhutzler.com/blog/2009/08/26/ ... he-get-go/

@maxblack,

Thank you for the tip!

I have remote root SSH access to the Pro unit and followed the directions in the link. Restarted RSYNC and this time it says it's doing an INCREMENTAL backup. It's building a list again, but it seems to be a different list than the one it was building before. Looks like it's working properly. :D

There are two hidden files in the /var/log/frontview/backup/ directory not mentioned in the link:

.last_full_backup_001
.last_full_backup_002

These files have the values:
1387747411
1389081378

Anyone know what these values mean and if I need be concerned about them?

Thanks again!