This blog is hopefully found before you really need it. If not, it's not uncommon to find that you've lost your files due to a disk failure, malware, or applications similar to crypto-wall. When an issue like this happens, you'll hear the same thing all the time "I wish I had a backup", or "Did you make a backup" and of course "I told you, you should have had a backup." Home users don't really need the fancy bells and whistles that others may need (like SoHo and SMB, or even Enterprise businesses). Although you won't really read that you need extensive equipment, and extravagant software -- what you can expect though is to obtain some tips to save your software, and maybe some tips on how to get your information back if a disk does fail.
External Hard Drives
You may read on a few web pages, and some magazines that you can purchase an external hard disk for a back up, and of course to "extend your storage." However, the main reason those hard disks are marketed, is for the sole purpose of extra storage. And there are reasons for this statement. First and foremost, external hard drives are exposed to more stress and shock (due to improper handling) due to their portability. Mix this with possibly having pets, or children which may increase the risks of the hard disks dropping, or sustaining spills, and it becomes a SINGLE point of failure. And, we will discuss the point of single VS multiple points of failure.
Considering we mentioned the first problem with the USB style hard disks, we shall continue onward with the next issue. Many users are not aware that these disks also need to be properly "shutdown" You shouldn't simply plug it in, and then rip out the plug. This can cause a few problems, and more than likely disrupt the data that is already on the hard disk that may be in use, waiting to be written, or even worse -- data that isn't being touched but may still be impacted. And, this is also said for USB Flash drives, as well as solid states, and you guess it platter based hard disks.
Internal Based RAID
Internal based RAID is a pretty good idea. However, due to the internal nature of the RAID / Backup it's not a true back up. Also, if in the event that the system is struck with a piece of malware like crypto-wall, and it's not caught in time, what can happen is a loss of data. Internal based RAID is also useless against theft, and we can also say that for all storage devices if they are not locked away. And, if they aren't locked away -- it will fall back on encryption. We will get to that later. The other issues that you may want to look into with internal based RAID similar hard drives may actually raise the failure rate of your hardware -- and what is worse without a true back up, your data may be lost forever. More so, if you don't have a say in building your own configuration or hardware, it will contain the same hard drive as the primary disk and that is exactly what most manufacturers do. So what this means is that if we have 2 hard disks that are Sea Gate, and they both have an uptime of 100,000 Hours and you've been using the computer and you're at your 99,999th hour the likelihood of both disks failing due to mechanical wear and tear are great. Instead dissimilar hard disks will benefit you in the long run. This way if one is rated for 100,000 hours and the other 200,000 hours, and either fails -- you still have a backup.
Adding a RAID controller to a P.C. is also asking for another point of failure. Should the controller stop working, or the drivers have problems -- this could spell a significant amount of data loss. In many instances, performance may be impacted because "true RAID" is usually never present within computers -- it's software based. More so it will not protect you against a virus, user error or worse -- deletion (intentional or unintentional). The point of this type of RAID or any RAID is to assist in a hardware (disk) failure. Many times disk RAID 0 is utilized for performance, but the problem with this is if you don't have 1 full working back up -- the chances of you recovering your data are very low.
Software RAID and, "Software RAID" (meaning most "hardware" based RAIDs on systems discussed above) have the same advantages of the previously discussed method but the problem with this aspect is that it's slower. If your system is built using a software RAID the performance will be impacted. This is mostly on the motherboard level, and is sometimes called FakeRAID. Normally to determine if you are on a SoftRAID one way you can tell is the price of your computer. Hardware RAID will normally cost a few hundred dollars more.
More so, software RAID is what you can do with an external hard disk, but that, too is not a true back up. The problem with these types of backup schematics is that either 1) No one has the time, 2) No one really remembers, 3) It takes too long, 4) Insert proper excuse here.
There are a few benefits of utilizing encryption with, RAID and that is of course theft. But the problem with encryption is that it will not only impact performance, it will also impact how the data is retrieved (and this is said with some strict eyes). Here is where management of a whole disk type of encryption schematic will be needed -- and you probably guessed it. Keeping track of that secret key you have going on over there. Although windows bitlocker does support hardware based raid -- the one thing that you'd need to do in other instances is with the stored hard disks (You know, your grandfather-father-son backups.) is to encrypt them before they are shelved. It may seem impractical to store these devices outside the home for a home user, but the suggested location would be a safety deposit box. If it's the security of the data that you are worried about, the best bet for you would be to have a static mounted device where the RAID will be.
If you do decide to encrypt your RAID be mindful that if in the event that the RAID set fails, you may need to deploy different tactics in order to retrieve your data. Again, remembering the key, pass-phrase, or other authentication method you've utilized. We've delt with a crash (where we had one disk that was still working) and wanted to validate the files on both hard drives however, Linux was giving us a headache. Figuring that the time spent to decrypt the files, validate them, and then re-install was a huge monetary loss we scrapped and fell back on one of the images we've made before the disk failed. If we didn't have the RAID set up with it's respective backup we would have lost everything!
Considering What to Backup?
Some people will say full disk is essential, others will say just the data. However, there are minor issues with each respective approach. Should you backup and archive your data, you'd need to install the operating system all over again. And, this can take time and energy (ugh -- all those programs!). If you perform a whole disk RAID you simply point the RAID to do it's thing and your back up and running within a short time. Of course many people would figure it would be a no-brainer but where is the real problem? Well... The storage.
So let's say you have a 1TB hard disk RAID, and the RAID is mirrored. That makes it a 1TB hard disk RAID. Cool. So you installed your operating system, you have a few programs -- a lot of pictures of the grand kids, videos, etc. In time what happens is that all those photos, and programs need to go somewhere, if the computer is not undergoing maintenance even the history can put you a little over the top. So now, what happens? You have 1.5TB of Junk, and 500GB of memories that you can't add anymore. So, here is where you have to decide whether you want full disk, or dedicated solution.
What is this dedicated solution you speak of? Well, if we are speaking of an internal solution for your systems, what you'd need to do is have 1 hard disk for the Operating system, your programs, settings, etc. Set up a secondary hard disk, and then a third hard disk. The second and third hard disks will be your storage locations. Whenever you save a picture disk 3 will mirror disk 2 and all is well and happy. Now, you're free to install all the software you want. The space will not be impacted by anything other than, more memories. Of course this can and will cost a bit more money because you're now purchasing 4 hard drives (1 for the os, 2 for the internal backup and of course one outside the computer for a backup of the internal RAID).
What Options are Available?
If you are really concerned with space the only logical thing you might be able to do is dedicate the efforts to an internal RAID. However, there are trade-offs as we have discussed. In this case you'd have a RAID and you can opt-in for the external hard disk that will backup your RAID just in case anything goes wrong. But you'd have to remember to backup and write your files to the external hard disk. Here is where it may become a bit dizzy regarding the setup options you'd have before you. The chart below will aim to break this down.
|3 Disk (Mirror + OS + External Backup)||Low|
|4 Disk RAID. 1 for OS 2 for data Mirror. 1 for External||High|
NAS (Network Attached Storage)
NAS has a lot of great points to it, and again it also has a lot of issues that you'd need to consider when attempting to go this route. First, you'd need to purchase a nas, and 3 hard drives (and this depends on your NAS -- We are speaking of such with the view of a 2 bay RAID + 1 External backup). Mostly these will be done through your switch, Firewall, or Router. The one thing that you should look into if your files are rather large is a gigabit connection for your devices (again, switch, firewall or Router). With this type of setup you can different folders "zones" that can assist you with backing up your data. You can set guest access, access to folders for the kids, and access for yourself if you should so wish to do so. There are two methods you can provide a backup to this type of setup. On-demand, and when needed. With on-demand your disks are always powered up, and always spinning. This is said 24/7 365. Although this can put strain on the mechanical aspects of your disks -- they can and will perform functions faster. And, by faster we mean not waiting for the disks to spin up and write the data. Although this may age the disks faster, in our environments we don't utilize this type of setup. The other portion you can use is with disk sleep. When the disks are not in use, or not needed the disks are spun down and await for a connection before waking them up.
You can use this type of setup only for your data, and the settings for your systems. If you require more storage space you can re-direct programs to utilize the NAS. When speaking of which programs will have access to the NAS we are referring to Microsoft Office, redirecting your OS pictures folder to the NAS so that when images are imported they are directly sent to the NAS, and other such important pieces of information. But of course you will need that secondary hard disk that will be the backup for the RAID. Now, you should realize that with this, you will still need to make an independent backup of your files on the RAID in case something goes wrong. This way you have multiple points of failure.
Points to Consider
- Disk RAID require the hard disks used for the RAID + 1 for additional backup.
- Should your RAID be out in the open, ACL or Encryption should be utilized.
- SoftRaid / FakeRAID impacts performance.
- External backup of your RAID should be encrypted.
- RAID does not protect against viruses, malware, user error, and or deletion.
- External hard drives, flash drives, CF/SD, etc. Are not real backup strategies.
- Some laptop computers come with 2, or 3 disk RAID.
- Consider backing up the full disk, or just data.
- There is no one size fits all!