Storage Reading Journal

Storage devices are devices that hold digital information and come in the form of optical media, flash media and magnetic media. A popular method for having data storage be available almost anywhere is cloud storage. Cloud storage is saving data across the internet to another location which houses the physical storage devices. Some Internet service providers and companies such as Amazon and Microsoft offer cloud storage services. These services could provide some amount of free storage before needing to pay for more space. These services also provide data backups.

Hard drives are one of the most popular storage devices and can be installed within a computer or attached to a computer via USB, IEEE 1394, eSATA, or eSATA port. External hard drives come in housings that may include fans. Another reason for the popularity of hard drives is their superior storage capacity compared to flash drives and faster data transfer when compared to tape drives. Hard drives are available in many different physical sizes, also known as form factors. 5.25 inch and 3.5 inch are used in desktops and smaller servers, the 2.5 inch was designed to be used in laptops and Ultrabook and the 1.8 inch is used for a wide variety of products such as SSDs, Ultrabook and easily portable devices.

Solid State Drives, more commonly known as SSDs, use either DRAM or nonvolatile flash memory to store data instead of the more mechanical storage method of hard drives. This allows SSDs to avoid the most prevalent cause of hard drive failure, the breakdown of moving parts. The modern SSD’s use of flash memory allows it to minimize heat production and noise while still allowing it to be fast, secure and long lasting. The SSD can be used both internally and externally similarly to hard drives but is the preferred storage device in tablets and other mobile devices. SSDs are also favored in environments that may have extreme temperatures or where the drive may be jolted. The differences between SSDs and hard drives does not prevent them from working together.

The way an SSD stores data is that it runs an erase operation, moves data to an alternate location and then actually records the information. Despite this process, SSDs still boast high performance capabilities. An important term is write amplification which is the minimum amount of memory storage space that is affected by a write request. This process does mean that writing takes longer than reading with SSDs. Wear leveling is when data is erased and written using all of the memory blocks in an SSD instead of using the same blocks repeatedly. The three methods of achieving this are using software to track and direct write operations, reserving a set amount of memory blocks to serve as backups when another block fails or using a combination of the two methods. SSDs use a NAND system to indicate where data is stored. A 1 bit represents no data is currently being stored in its particular location and a 0 bit shows that there is data. NAND flash memory allows for data to be preserved when the device loses power. Two types of technology used in SSDs are single level memory cell(SLC) and multi level memory cell(MLC). With SLC each memory cell gets 1 bit stored in it. SLCs last longer than MLCs but are more expensive. MLCs are able to store more than 1 bit per memory cell and are cheaper but are slower in transfer speeds.

The main con of SSDs are that they are expensive when compared to the price of hard drives. SSDs have a limited number of reads and writes per memory block. SSDs that use their entire memory capacity last longer. Solid state hybrid drives(SSHD) are a combination of the mechanical technology of a hard drive and the flash technology of a SSD. Data that is more frequently used is stored on the flash memory for quicker retrieval while lesser used data is sent to the mechanical storage. SSHDs provides a lower cost compared to an SSD while still having some fast memory storage while having a large amount of storage space provided by the mechanical drive. A piece of technology that is similar to a SSHD is the flash cache module(FCM). The FCM uses software to predict what data is going to be needed and then transfer that data to an SSD that is a separate device than the mechanical hard drive.

Hard drives use sets of rules to operate. Some of these rules specify how many heads the driver has, the amount of devices it can support and how many data bits can be transferred at a time. These sets of rules are called an interface and they control communication with the hard drive. The two main types of interfaces are the integrated drive electronics(IDE), which is also known as the AT Attachment(ATA) or Enhanced IDE(EIDE), and the Small Computer System Interface(SCSI).IDEs are most commonly used in home and office computers while SCSIs are used more in network servers. There are other interfaces that are used with external storage devices such as USB and Apple’s Thunderbolt. Both IDE and SCSI started out using parallel architectures which meant that multiple bits would be sent out over multiple which created the need for very precise timing as transfer rates became faster. Both interfaces allowed for multiple devices to attach to the same bus but Parallel IDE only allowed two devices while Parallel SCSI allowed more than two. The idea was the same for both interfaces, if multiple devices are using the same bus, then they have to wait for their turn to arrive. Modern interfaces are moving more toward serial architectures which is a point to point bus which allows for each device to have its own individual connection that goes to the controller where bits are sent one at a time. This allows for easier scalability. The ATA that uses a serial architecture is known as a Serial ATA(SATA) and SCSI is known as a Serial Attached SCSI(SAS). SATA in laptops is being replaced by the M.2 interface which allows for modules of a variety of sizes to be attached.

Cluster: The smallest amount of space reserved for one file. Made up of a specific number of sectors. Sector: A section of a track. Usually 512 bytes. Cylinder: The same track across all platters.

A drive array is when two or more hard drives are configured together for speed, redundancy or both. A redundant array of independent disks(RAID) makes it possible for reading from and writing to multiple hard drives in order to have larger storage, better performance and an increase in fault tolerance. Fault tolerance is how capable a system is able to continue running in the event of a hardware or software failure. RAID can be used with both hardware or software with the hardware RAID being configured through BIOS while software RAID would be configured either through windows or a software provided by the manufacturer. RAID can also be used on flash cache modules(FCMs) and mechanical hard drives. RAID comes in levels and levels 0,1 and 5 are available for windows. RAID drives are often made so that the can be removed and replaced while the computer has power, this is known as hot swapping. RAID use to require an adapter and software to work but many modern motherboards naturally support RAID. Modern OS also support RAID but the BIOS/UEFI will most likely have to be configured.