Computer System Architecture - Finals (Module 1: Computer Storage)

Serial access

This means that when reading a data item, all previous items have to be read or skipped. The major limitation of this storage devices is that the location of the searched item influences the search time.

Random access

It is possible to get to the required item directly. Memory, for example, is a direct-access device. The access time for obtaining a data item using direct access storage devices is almost similar for all items, regardless of their location.

Primary Storage

The CPU accesses the data or instruction in the conventional memory (RAM) if the data in the cache memory is not present. Both conventional and cache memory provide immediate access to program instructions and data by the CPU and can be used for the execution of programs.

Secondary Storage

This is the level of conventional memory in which the storage in the hierarchy is not immediately available to the CPU. This is treated as the I/O where data and programs must be copied to primary memory for CPU access. Except for flash memory, access to secondary storage is significantly slower than primary storage.

Magnetic Disks

These consist of one (1) or more flat, circular platters made of glass, metal, or plastic, and are coated with a magnetic substance. Particles within a small area of the magnetic substance can be polarized magnetically in one of the two (2) directions with an electromagnet. An electromagnet can also detect the direction of polarization previously recorded. Thus, magnetic polarization can be used to distinguish 1's and 0's.

track

With the head in a particular position, it traces out a circle on the disk surface as the disk rotates; this circle is known as a ________

cylinder

Since the heads on each surface all line up, the set of tracks for all the surfaces form a __________.

sectors

Each track contains one (1) or more blocks of data. Traditionally, the surface of the disk platter are viewed as divided into equally sized pie shape segments, known as _________, although current disks divide up the track somewhat differently.

Access time

This refers to the time that takes to start transfer data. It is also the sum of seek time and rotational delay.

Reading or writing time

This refers to the time required for the head to read or write the data. The reading or writing time depends on the electronic data transfer (the head is over the requested location, and all it has to do is read or write the data).

Seek Time

This refers to the time that takes the head move to the proper track.

Rotational Delay

This refers the time that it takes a sector to be positioned under read/write head, depending on rotation speed.

Rotation speed

This is represented by revolutions per minute (RPM) and assumes that the sector is away from head half of the track; therefore, the Disk File System calculates the rotation delay.

File system

This defines the organization of information stored in the hard disk.

Solid State Memory

It is a solid state storage medium that uses volatile memory as a means of holding and accessing data. It is based on NAND Flash chips which differs from RAM in terms of usage and volatility.

Flash memory

is well suited for secondary storage in many cases. Because of its small size, low power consumption, and light weight, it is frequently the secondary storage of choice, which is used for both long-term system storage and portable storage.

Optical Disks

These are an alternative to magnetic disk storage. These include various types of CDs and DVDs, in read-only, write-once, and read/write forms.

Magnetic Tapes

Like other magnetic media, tape is nonvolatile, and, under proper storage conditions, the data can be stored indefinitely. Generally, full system backups are made to tape and are moved to off-site locations for long-term storage. Tape is stored in a cartridge using a standard format called LTO (linear tape open).

When the tape is in the tape drive ready for operation, it is said to be mounted. Tape cartridges have the major advantage of convenience. They are easy to mount and dismount, and small and easy to store. Current tape cartridges can store as much as 3 TB of compressed data or 1.5 TB of uncompressed data. Cartridges with uncompressed capacities as large as 12 TB are currently in development.

The technique used for this type of storage and retrieval is called data streaming.

Parallel Advanced Technology Attachment (PATA)

These were the first types of hard disk drives that made use of Parallel interface standard to connect to computers.

These types of drives are referred to as the Integrated Drive Electronics (IDE) and Enhanced Integrated Drive Electronics (EIDE) drives.

These drives store data by the use of magnetism. The internal structure is made of mechanical moving parts.

Western Digital introduced this drives back in 1986. Data transfer rate can go up to 133MB/s, and a maximum of two (2) devices can be connected to a drive channel. Most of the motherboards have a provision of two (2) channels, thus can connect a total of four (4) EIDE devices internally. this drives make use of 40- or 80-wire ribbon cable to transfer multiple bits of data simultaneously in parallel.

Serial Advanced Technology Attachment (SATA)

These hard drives have replaced the PATA drives in desktop and laptop computers. The main physical difference between the two is the interface, although their method of connecting to a computer is the same.

Here are some of the advantages of _______ hard disk drives over PATA:

o It can transfer data faster by using serial signaling

technology.

o The cables are thinner and more flexible than PATA

cables.

o It has a 7-pin data connection with cable limit of 1 meter.

o It does not share bandwidth because there is only one (1)

disk drive allowed per ______ controller chip on the

computer motherboard.

o It consumes less power. It only requires 250 mV as

opposed to PATA's 5V.

Small Computer System Interface (SCSI)

These are quite similar to IDE hard drives, but these make use of parallel interface standards developed by the American National Standards Insitute (ANSI) for attaching devices to computers.

_____ drives can be connected internally or externally. Devices that are connected in an _____ can be daisy-chained and has to be terminated at the end through the use of a special terminating plug designed for ______ devices.

_____ can support devices like multi-hard disk, CD-ROM, and scanner. It can also connect as many as 30 devices and peripherals by daisy chain. It is the most suitable for an entity that needs enormous capacity.

Here are some of their advantages:

o Faster and very reliable.

o Good for 24/7 operations

o Have better scalability and flexibility in arrays

o Well-adapted for storing and moving large amounts of

data.

Serial Attached SCSI (SAS)

It is a new technology of hard disk attachment. It combines the advantages of current parallel SCSI and serial attached technology, sets serial communication as the protocol infrastructure, adopts SCSI-3 extended instruction set, and is compatible with SATA devices. It is a multi-level storage devices stack; in fact, this disk is a disk adopting this interface technology.

Here are its characteristics:

Better Performance - It reduces the address conflicts and the deceleration of daisy chain. It provides a special signal channel to each device to ensure the maximum bandwidth; Data operation under full duplex ensures the most effective data handling capacity.

Simple Cable Link - Thinner cable matches smaller connector.

Better Scalability - It can connect more disk devices simultaneously.

Larger blocks

The controller can read larger quantities of data from the disk as compared with the amount requested by the application. The additional data is stored in buffers within the controller.

Pre-fetch

The controller can identify the application's patterns of behavior. When an application reads the sectors sequentially, the controller can read the next blocks even before they are requested.

Access algorithms

These are usually implemented by the operating system. These are included in the discussion to provide an elaborated explanation and a complete view of performance enhancement methods.

Disk array

This is a technology that increases the size of the single disk as well as its speed, reliability, and survivability.

Storage Access Algorithms

There are various algorithms that are implemented by the operating system to optimize the disk access time.

First Come First Served (FCFS)

handles the requests based on their arrival order. The first request will be served first, followed by the second request, and so forth.

Shortest Time First (STF)

handles the requests based on their distance from the current head location. The first request to be handled is the closest one.

SCAN (elevator algorithm)

is where the head starts at one (1) end of the disk (usually at the first cylinder) and, while moving to the other end, handles all the requests found along the way.

When the head reaches the end of the disk, it starts its movement backward following the same mechanism, only this time the sorted numbers are handled from the largest downward.

Circular SCAN (C-SCAN)

is very similar to SCAN but with a minor change. The head starts at one (1) end (usually at the beginning of the disk) and, while moving to the end of the disk, handles all the requests it finds.

C-LOOK

is an improvement of C-SCAN. Unlike C-SCAN, which moves from the beginning of the disk to its end (even if there are no additional requests near the end), the head moves forward until it finds the request with the larger cylinder number.

Disk Arrays

can be used to reduce the overall data access time by sharing the data among multiple disks. It can also be used to increase system reliability by providing storage redundancy.

One useful type of this is RAID (Redundant Array of Inexpensive Disks).

Mirrored array

It consists of two (2) or more disk drives.

In this, each disk stores exactly the same data. During reads, alternate blocks of the data are read from different drives and then combined to reassemble the original data. Thus, the access time for a multiblock read is reduced approximately by a factor equal to the number of disk drives in the array.

Striped array

It uses a slightly different approach in which a file segment to be stored is divided into blocks. Different blocks are then written simultaneously to different disks. This effectively multiplies the throughput rate by the number of data disks in the array.

It also requires a minimum of three (3) disk drives; in the simplest configuration, one (1) disk drive is reserved for error checking.

Cloud storage

refers to remote data storage done over a computer network. Instead of using an external device connected to the computer, cloud storage makes use of storage devices located elsewhere and connected by the Internet or some other type of network connection.

Storage Attached Network (SAN)

It is a local area network intended to handle large amounts of data and provide fast transfer rates to organizational data.

One of the advantages of using this is that it provides the user community with simple, coherent access to common corporate information.

Network Attached Storage (NAS)

It represents a different approach to storing data in the network. In this, the storage resources are connected directly to the network, unlike the other one where the storage resources are part of the network.

This approach has been around for a long time as a method for sharing files. Contrary to the other approach, which allows access only on a block basis, this relates to the whole files.

Better Scalability

It can connect more disk devices simultaneously

Simple Cable Link

Thinner cable matches smaller connector.

Better Performance

It reduces the address conflicts and the deceleration of daisy chain. It provides a special signal channel to each device to ensure the maximum bandwidth; Data operation under full duplex ensures the most effective data handling capacity.