Data Protection Proposal
Lake Medical Hospital

By: Frank Angelone
05/31/14

Lake Medical Services in this proposal I will inform you of the options. The Standard for using data protection is by using RAID. RAID stands for Redundancy Array of Independent Disks. What this system allows you to do from an owner’s perspective, is to store and process all of your network needs. Considering you are a hospital you need to protect your patient’s information. With this data protection system it will allow you to collect large amounts of data by combining hard drives into one. What the RAID system does is take many hard drives and appear in your network as one. There are three main standards of RAID. First is R.A.I.D. 0, second is R.A.I.D 1, and third is R.A.I.D. There is a fourth option we will discuss that I am proposing for you to use. If you would like another option this can be discussed, but first we will discuss the three main standards.
First option of R.A.I.D. is 0. When using option 0 you need to have at least 2 hard drives. You may use more drives than to increase the speed of your network (Iomega, 2009). Also looking at the size of the network will determine how many drives will be needed to run the system. This option uses a method called striping (Iomega, 2009). Striping is when the data that is stored on the drive is spread across the disks on the drives. What this means is that every piece of data is split up into small segments. Then these small segments are spread out through the disks on either drives or even multiple drives depending on how many hard drives you have. Looking at the diagram will give you a better understanding of how this system stripes data across the drives.

So if you save a word document a piece of the document would be saved on Drive A Disk A1, and another piece would be stored on Drive B Disk1. Remember even though I show a drive A, and drive B both of these drives act as one drive on the network (Iomega, 2009). So the document gets stored on the one drive of the network, but is broken up among the disks on each drive. Because the document is broken up into segments the system will process faster because both drives are sending out data to the user (Iomega, 2009). So with this particular system you achieve great speed, but there is no protection. If one drive fails you lose all of the data. Because it breaks up into segments each drive has only a piece of the puzzle. If a puzzle piece is missing the puzzle can be completed. So there is no protection with this system but provides a lot of speed. Now I don’t recommend this particular system for you but later another option has a similar feature.
The second standard that is used with R.A.I.D. is R.A.I.D 1. This system has to have a minimum of two hard drives (Iomega, 2009). The method that this standard uses is mirroring (Iomega, 2009). Yes it operates exactly as it sounds. The two drives still operate as one. This time when you download data to the hard drive it is automatically downloaded to the other hard drive. The major benefit of this is that when one hard drive fails you can replace the bad drive while the system is running. This system is all about redundancy. It allows the server to maintain its operation. However this system has its limitations. Because of the mirroring you have to size your network, then after knowing how much gigabytes you need. You will have to ensure that each drive has enough gigabytes to run the entire network. The reason for this is because one drive acts as the server and the other is a backup. But if one drive does fails, when you replace the bad hard drive the system backs up automatically. Because of the mirroring aspect it can make the server run slow depending how you size your hard drive.
The third standard of R.A.I.D. is 5. This is more of an option in which your company is looking forth. By looking at the other two options one work well with speed and the other works well with redundancy. Well this standard takes both concepts and combines them. This system uses the striping method that we talked about with standard 0. But this standard needs a minimum of three servers in order to function (Iomega, 2009). The reasoning is that not only does it use stripping, it creates redundancy that allows one server to go down so that you won’t lose your data. Looking below at the diagram may help you understand.

So if you look at each drive. Each color represents one document. This diagram is not exactly how it is duplicated but it is provided to give you the idea of how the concept works. When the document is split up into segment the segment is sent to two different drives. Then if a one drive goes down then the other drive had the same segment so the data isn’t lost like with the standard 0. So looking at the diagram above, it takes two colors to maintain the data. So if you lose one drive you still have both segments of the data they call this parity (Iomega, 2009). So this is how R.A.I.D. created protection while increasing speed. To understand how much you need to size the RAM you have is determined by the combination of two of the drives not three. This is because the third drive is merely created for redundancy protection. Also when combining different size drives you can only function at the rating of the smallest drive (Iomega, 2009). So if you have more than three drives you would add up all of the RAM then subtract one. So if you lose one drive the server still operates while you replace the down hard drive.
So now that you understand the three basic standards of R.A.I.D. I am going to prose to you that you use R.A.I.D. 10. With this system it can be costly but you would have to have almost half of your hard drives down before your server would go down. Because you rely so heavily on your network not to fail you, this selection is going to be the best suitable. R.A.I.D 10 combines standards 1 and 5. What it does is mirrors two standard 5’s. So not only do you have striping with parity but you have mirroring going on as well. So if you only used six hard drives four of the hard drives would have two fail and two on each mirroring set would have to fail. Also this system allows you to replace hard drives while it still is running.
Now let’s talk about the operating system. You can go software because you do have a system currently in place. But because of the problems you explained to me about the failing of your server, I don’t recommend the software option. If you chose the software option the hard drive the software is on if it fails you will lose the system. So I propose you use the Hardware drive R.A.I.D. controller. This method allows the system to function the way you need it. Because it is hardware if it goes down the operating functions as if one of your hard drives went down. While the server is running you can go in and replace the hard drive and the system can be rebuilt. Because of the size of your network system and the importance of your network having R.A.I.D. 10 is a good option for you. However we strongly suggest having a spare R.A.I.D. controller on hand for critical emergencies. You may never need the controller but this is a product that isn’t sold on most suppliers shelves and has to be ordered.
Let’s talk about your natural disaster plan. Because you have generators in place emergency power is available. So this power service needs to power up your servers. Plus using surge protectors on the power circuit for these drives is recommended. Also the room that these servers are in will need to have an independent cooling source to keep servers from overheating. Also its own fire suppression system needs to be in place to protect against fires. As well as the room located in a place where the structure is supported well in case of an earth quake. Considering your structure has been planned for natural disasters these task should be easily obtained.
So this is the system we recommend for you. It provides the redundancy that you need. As well as keeping the servers up at all times. If the server would to crash it would not take very long for you to get the server back online so your hospital could function even under the worse circumstances.

References
Iomega, 2009: retrieved from. http://iomega.com/resources/pdf/pdf_166.pdf