Reading Journal ‐ Storage - sullivaneg/TechJournal-SYS140 GitHub Wiki

Storage Reading Journal

Storage Devices Overview In this section I learned that data can be stored in physical devices such as optical, flash, and magnetic media. I also learned that data can be stored “in the cloud” meaning that it’s stored on servers that companies own. You can rent space from those companies and store your data there. This section was fairly easy to understand. I’ve used different types of physical storage such as flash drives, SSDs, and cloud storage so this overview was review for me. Hard Drive Overview In this section I learned about hard drives and more specifically magnetic hard drives. I learned that hard drives can be inside the computer or attached by USB. They store a lot of data, more than flash drives. Hard drives can come in different sizes, these are called forma factors. Typically larger hard drives are used for desktops and smaller ones are used for laptops. I learned that most traditional mechanical drives are magnetic. They have small metal parts called platters. Each platter holds data on two sides and has two heads that read/write. The data in the drive is written with electromagnetism. Magnetic hard drives usually have two motors, one is used for the platters the other moves the read/write heads. These drives spin at different speeds measured in revolutions per minute (RPM). The faster the hard drive spins the faster the data can be transferred and the more expensive the drive is. I learned that these physical drives can often fail. One way that they fail is a head crash, this is when the read/write head hits the platter. This can cause corrupted data. MTBF or Mean time between failures is the average number of hours before a drive might fail. This section was interesting because I’ve always been fascinated by the way that discs store information. Like CDs or DVDs and hard drives are just another version of this electromagnetic way of storing information on a spinning disk. I would recommend this section because it has good diagrams that really help break down the cylinders and the tracks, and the sectors within them.

SSD Overview In this section I learned about Solid State Drives or SSDs. SSDs are storage devices. Newer ones use flash memory, and older ones use DRAM. SSDs are interesting because they get rid of the main reason hard drives fail, they have no moving parts. They use flash memory usually so they’re more reliable and fast. They are used in laptops, desktops, tablets, and some mobile devices. They are used in a lot of industries where the drive might be disturbed or you might encounter extreme temperatures, such as medical, industrial and automotive industries. SSDs differ from hard drives in the way they are written too. Two terms used when describing the writing process in SSDs are write amplification and wear leveling. Write amplification is the smallest amount of memory space that is affected by writing. SSDs sometimes need to erase data, move data somewhere else, and then write. So the write amplification is the smallest amount of data that would be affected by that process. Wear leveling is a technique that is used to erase/write data by using different memory blocks instead of the same ones over and over again. SSDs use 1s and 0s to figure out whether there is data stored in a location. There's a 1 bit when there is no data and a 0 bit when there is data. I also learned that SSDs use two types of technology, single-level memory cell (SLC) and multi-level memory cell (MLC). SLCs keep 1 bit in each memory cell. MLCs keep over 1 bit in each memory cell. MLCs are cheaper but slower. This section was interesting because I personally use an SSD for storing photos from my camera since I shoot in RAW and they take up a lot of space until they are edited and converted to JPEG. I got this SSD like 4 or 5 years ago so I didn’t really understand how it worked or what kind of SSD to look for. This is interesting because now I can look up that SSDs specs and learn about what I’m using and how I can better store my data in case my SSD fails.

Hard Drive Interfaces Overview In this section I learned that hard drives have a set of rules about their operation called an interface. These standards cover the number of heads, commands, number of devices supported, number of bits of data transferred at one time, etc. There are two main types of interfaces. Integrated drive electronics (IDE) and Small Computer System Interface (SCSI). IDE is common in home computers and SCSI is used in network servers. IDE and SCSI are parallel architectures, meaning that multiple paths are used to send multiple bits of data. This section wasn’t super interesting to me but it was cool to learn what those pieces of hardware were because I encountered some in my stint of taking computers apart and I think my little kit of hardware for my raspberry pi might have come with something similar. So that was interesting.

Hard Drive Preparation overview In this section I learned that once a hard drive is installed, it enters a setup program. There are two steps in the hard drive preparation, step one is to partition the drive and step two is to high-level format the drive. Partitioning the hard drive will let a drive letter be given to 1+ parts of the drive. It lets the computer see the drive as more than one drive. Originally you would partition drives because then you could run multiple operating systems. Partitioning also allows more efficient use of space. High-level formatting prepares the drive to be able to use a specific file system. A file system is a system used to store data on the drive. Windows commonly uses FAT16, FAT32, exFAT, and NTFS. A drive can’t be used until it’s been partitioned and formatted. In this section I also learned about what a cluster is. A cluster is the smallest amount of space for one file and it’s made up of sectors. I also learned that when you partition a drive you can only have one part be the primary partition. The other cylinders can be extended partitions. The extended partitions allow the drive to divide up into logical drives or volumes. There can be only one extended partition as well. This section was interesting because I read this section after we did the lab where we actually got to use diskpart and practice dividing drives into different sections. So I actually understood part of what this reading was talking about and it was cool to see diagrams of the physical representation of what we were doing during the lab.

Fault Tolerance In this section I learned about how drives use redundancy to minimize the risk of losing data. I learned that a disk array is the use of 2+ hard drives that are configured to be fast or redundant or some combination of both. Redundant array of independent disks (RAID) reads and writes data from and to multiple hard drives to protect data or increase the speed of the writing process. There are 5 levels of RAID. RAID 0 or “disk striping” alternates the writing between two or more drives. This makes the writing of the data more efficient but doesn’t protect data because if one drive fails you lose data. RAID 1 or “disk mirroring” is when you use 2 or more hard drives and a disk controller. The data is written to two drives, that way if one drive fails the data is still in other drives. This protects against failure but is slower because you’re writing everything twice. RAID 0+1 is a striped and mirrored set put together. You need at least 4 hard drives and you need an even number of drives. This way creates a second striped set to mirror the original striped set. RAID 1+0 or RAID 10 is also a mirrored and striped set combined and you need at least 4 hard drives. 1+0 has a striped set made from a mirrored set. The last RAID type is RAID 5 is also “disk striping” but with parity. It writes data to 3+ hard drives. There’s parity included in the data so if a drive fails the data can be made again from the information in the other two drives. This section was also interesting because I read it after I did the lab. I actually understand how at least the striping and mirroring works. It makes sense to me. It’s also making me wonder if I should set up a more redundant form of storage for my files because everything of mine is stored on one form of storage and now I’m concerned about them failing.