Back Up

[friendlyguy]

One is continually reminded that RAID in any configuration is not a back up. I thought that was the whole purpose of RAID 1 when it was introduced--an always up-to-date back up.
Unfortunately, to store 6 TB (2x3TB for example) required 12 TB of storage media. So.. along came RAID 5. Redundancy with just another 3TB drive. Then along came RAID 6 with double redundancy and just one more drive. BUT....we're still told we need 2 more 3TB sitting somewhere (firebox for example) "just in case". Well, I understand the flood, fire, and theft risks but these other back ups are not likely getting updated very frequently. Furthermore, they are subject to bit rot. Why not just stick to RAID 1.
1. Why have RAID at all if you already have a back-up somewhere? Just use JOBD in your enclosure.
2. What about just another 2 drives in the same enclosure as a "second back up" (2 with media, 2 as back up and a further 2 as back up for those who are paranoid)?
3. Snapshot we're told, isn't a back up either but merely a restore point, although is all an out of date back up would give. How many disks (space) does Snapshot require, relative to the basic storage. Would one need 6 TB for Snapshot to match the original 6 TB?
Thanks

[raulfg3]

How many disks (space) does Snapshot require, relative to the basic storage. Would one need 6 TB for Snapshot to match the original 6 TB?

0=cero=none , please google a bit about how snapshot works in ZFS.

Why have RAID at all if you already have a back-up somewhere? Just use JOBD in your enclosure.

ok, no problem with this, but back-up somewhere MUST not be in the same case that original data

[crowi]

A backup saves you also for accidentally data deletion and not backupped data can be considered of lost data.
Never ever put actual data and backup in the same place, as Raulfg already mentioned.

[friendlyguy]

Thanks for the replies.

[achaycock]

It's certainly true that RAID when used for redundancy purposes is not a backup, but it is useful for resilience to failure. What this means is that a drive can fail without the whole system/share going down.

If we take the JBOD scenario as an example, supposing you have a decent backup solution but one of the drives fail during the course of a backup? Any new data added between the last and current backup will be lost, not to mention the unavailability of the storage until such time as you can source a replacement drive, rebuild the JBOD array and restore the data that you do have. Now this isn't to say that using JBOD is wrong or that the risk is necessarily a bad one. It may be that you can afford the downtime far more readily than you can afford the cost of redundancy, but I will add that in my personal experience I have found that RAID 1 has saved my bacon on a few occasions now with systems continuing to do their job despite a drive dying.

In my view backups are essential, RAID is not a backup, RAID is worth the money due to the time it save.
In my personal experience, I have found that RAID 1 has frequently been enough to make it unnecessary to call on a backup - though I stress again that a backup IS vital.

[friendlyguy]

It's just a home server, so it doesn't matter whether its down for a while. Yes, I understand that anything that was saved after the last back up would be lost.---and is an important consideration. Thank you for your advice.

[Joyce J]

If you are looking for simple way to backup then try Ahsay Software. Ahsay Backup software is the best backup solution has it all real-time archiving, unlimited versions of files, fully automatic and it designed for manage service provider to provide online backup service to their users. I’m really happy with it. You can also try this.