Mitigating the Effects of Data Loss : RAID Vs Backups
Friday November 18, 2016
What is the value of a server? In this world of virtual machines and dedicated servers, our customers are becoming more and more removed from the physical components that comprise a server. Everything is easily replaceable — everything except the data contained within the servers.
Countless work hours have gone into making each server unique, with custom set-ups, modified WordPress templates, blog posts going back years, etc.
This is where the value of a server lies, in the data. What is this data worth to you? How do you even begin to measure that?
The data is stored on the server’s hard drives.
And guess what? This is, by far, the most common part of a server to experience failure. So it’s absolutely necessary to create a backup strategy.
So how do you to create the best backup strategy? Where do you begin?
There are two ways to prevent against the effects of data loss from drive failure – prevention and recovery.
On the prevention vector, we focus on RAID: Redundant Array of Independent Disksvarious
RAID configurations can be implemented to allow your data to withstand the loss of one, two, or more drives. There are several different configurations that can be tailored to your specific needs, essentially finding the sweet spot between performance, resilience, and cost that is right for your environment. RAID uses two or more drives to store your data in ways that can not only survive drive loss but can also improve performance.
However, even RAID can only protect from so many simultaneous failures. While it certainly helps prevent data loss in most cases, it doesn’t reduce it to statistic nothingness. Servers are susceptible to multi-drive failure, which is more common than one would expect.
When setting up a server with multiple drives, often these drives are all from the same batch, if they are installed new. If one drive has a flaw, it is likely that this flaw is shared by the other drives in that batch, making the loss of 2 or more drives in a short time more likely than one would expect. In addition to often coming from the same batch, drives in a server are exposed to the same environmental factors, as well.
Furthermore, in addition to the fear of multiple drive failures, files and filesystems can become corrupt — either accidentally or maliciously. In this case, no RAID level will help you out of this jam.
This is where our second vector comes into play – recovery.
Backing up your data to an external system and keeping multiple recovery points is one of the best ways to mitigate the effect of unpreventable data loss. No matter how robust your data storage plan is, it can fail.
Our R1Soft backup service is set up to take daily incremental backup of your data.
Your data got corrupted on Tuesday? No Problem! Just restore the data from Monday.
You lost the “wrong two” drives in your RAID 10 array? Easy! We’ll simply replace the bad drives, and any others that don’t test 100% stable, then perform a bare metal recovery of your OS and data.
Every client with the service has full control and visibility of their backups with the ability to review, edit and download their backups using a personalized interface.
So, should you choose RAID or R1Soft backups?
While RAID alone is a great option that can prevent the downtime associated with recovery on single or even multiple-drive failures (as long as they are the “right” drives that fail, but that is another topic), it is not failproof.
On the other hand, backups alone can require lengthy downtime for recovery to occur, and the backup is only as current as the last recovery point.
This is why, for protecting your cannot-lose data, we recommend the dual-vectored approach of prevention and recovery. Significantly reduce the need to recover from backups by using a robust RAID, but have those backups on hand for when you do need them.
If you would like to add RAID or R1Soft backups (or better yet, both) to your current setup, chat with our specialists.