Cloud Computing Fundamentals

International & Access Foundation Programmes

• This is a computer science module
• The module involves computer networking, PC internals, storage and network storage
• The semester is semester 1
• The module part is part 6

Computer Hardware

• A computer's hardware structure includes input (keyboard, mouse), processing (processor chip), and output (monitor, printer).
• Storage devices include hard drives and flash memory.

Inside a PC

• A computer's internal components include the CPU, RAM, CPU Cooler, Graphics Card, Motherboard, 5.25" Bays, 3.5" Bays, and PSU(power supply unit), Hard Drive.

Motherboard Components

• The motherboard contains DRAM memory slots, power connectors, CPU heat sinks, mounting points for fans, and a CPU socket.
• Other components include the Northbridge (with heat sink), Southbridge, ROM, PCI slots, hard drive slots, a backup battery, and AGP slot, connectors for peripherals (e.g., audio, Ethernet, serial port, USB, etc.).

DDR SDRAM Versions

• Different DDR SDRAM versions (DDR, DDR2, DDR3, DDR4, DDR5) have varying release dates, bus clock speeds (MHz), voltages (volts), transfer rates (MT/s), and prefetch burst times (in minutes).
• DDR versions are characterized by their transfer rates and speeds.

What is RAM?

• RAM (Random Access Memory) is computer data storage used for programs currently in use.
• Most RAM chips lose their memory when the computer is powered off, meaning data needs to be loaded from storage each time a program is opened.

What is DDR SDRAM?

• DDR SDRAM is a type of RAM.
• DDR SDRAM stands for Double Data Rate Synchronous Dynamic Random-Access Memory.
• It has seen many iterations since 2000. Recent versions use higher transfer rates up to 6400 Mbps with capacities up to 32GB.

CPU Components

• The CPU has an ALU (Arithmetic Logic Unit) for calculations.
• Registers are internal CPU memories that temporarily store data during processing.
• Buses (Data Bus, Address Bus, Control Bus, Power Bus) are pathways that transfer information between the CPU, RAM, and other components.

Chipsets

• A chipset manages data flow between the processor, memory and peripherals.
• It is usually located on the motherboard.
• Chipsets are designed and matched to specific microprocessor families.
• They determine system performance.

Memory Addressing

• 32-bit systems can access 4GB of RAM.
• 64-bit systems can access much more RAM, making large programs possible.
• The speed of calculations per second affects the speed at which tasks complete in a given processor type.

Paging

• Paging allows a system to simulate more memory than physically exists.
• System loads data and program instructions into RAM, while swapping out less used data to the hard disk.
• This minimizes memory usage and allows running programs with more data than currently in RAM.

Swap File

• A swap file acts as virtual RAM, using secondary storage for data.
• The system moves inactive data blocks to the hard disk to free up RAM for use by active programs.
• This swap is used when RAM is full.

CPU Cache

• A CPU cache is a small memory location within the CPU that stores recently accessed data.
• This speeds up data access to main memory.
• CPU caches check the fastest level cache (L1) first. If data is in cache, it is accessed faster.

CPU Cache Levels

• When a processor needs information from memory, it first checks the cache.
• If in cache, it is much faster to process than going to main memory.
• Hit rates of caches can change, based on the amount of data in the cache.

Video Components

• VGA connectors connect video cards for data transmission.
• RAMDAC converts a digital image signal into an analog signal for the monitor.
• Video Socket inputs video from VCR analog signals.
• SVGA connectors connect the graphics card to a monitor.
• Video BIOS contains graphics mode definitions and screen fonts.
• Graphics Processor chips in video cards control elements associated with displaying images.

BIOS

• POST (Power On Self Test) is pre-operating system testing performed when starting the system.
• Bootstrap Loader prepares the system for operation for the operating system.
• BIOS Drivers run the system's drivers.
• BIOS Setup is for setting the system settings.

Types of Disk Storage

• Mechanical hard disk drives (HDDs), which store data using magnetic polarity, are the traditional storage system.
• Solid-state drives (SSDs) use flash memory and have much faster access speeds since there aren't moving parts.
• Flash drives use flash memory and are widely used for portable storage.
• Optical drives (CD-DVDs) use lasers to store data. They are less common now.
• Tape drives are less common today, more so for large-scale data backups.

Mechanical Hard Disk Drives

• The physical structure of a mechanical hard disk drive consists of a platter, spindle, head, actuator arm, actuator axis, power connector, and jumper block, IDE connector.

Hard Disk Drives Data

• Hard disk drives use cylinders, tracks, and sectors to organize data. Data is organized on the platter in concentric circles (tracks) and stacks of platters.

Solid State Disk (SSD)

• An SSD has a cache, controller, and NAND flash memory.

HDD and SSD Internal Architecture

• HDDs have rotating platters, a spindle, a head, and an actuator
• SSDs have NAND flash memory, a controller, and DRAM

Solid State Disk (SSD)

• Access times are much faster for SSDs than HDDs. A typical SSD access time is about 35 to 100 microseconds versus 5000 to 10,000 microseconds for an HDD.
• SSDs are more expensive than HDDs, particularly the larger ones.
• Reliability is typically better for SSDs versus HDDs. They generally have no moving parts, making them less prone to mechanical failure.
• HDDs are less expensive than SSDs.

Optical Storage

• Optical storage, like CDs and DVDs, uses lasers to read and write data.
• Data is stored in a pit pattern under the surface.

Making Disk Storage Reliable

• Backup regularly,
• Use RAID (Redundant Arrays of Independent Disks) to create a redundant array of disks to improve reliability.

RAID Levels

• RAID 0 uses striping to divide data across multiple disks for increased performance.
• RAID 1 uses mirroring to create a duplicate copy of data on multiple disks.
• RAID 10 combines mirroring and striping for good performance with a higher degree of redundancy and reliability.

RAID Arrays in a Server Rack

• Real-world RAID arrays are typically found in server racks, which contain storage disks formatted with RAID arrays.

Blade Servers

• Blade servers are stripped-down servers designed to minimize space usage and increase efficiency.
• They usually use a type of enterprise storage system like SAN (Storage Area Network) or NAS (Network Attached Storage) to support them.

File Systems

• File Allocation Table (FAT) and New Technology File System (NTFS) are methods for organizing data in file systems.
• NTFS is the newer, more advanced file system in modern operating systems. It's used in recent versions of Windows, and also has support for Linux and other operating systems.
• FAT is used for older storage systems, most commonly used in older Windows operating systems.

Disk FAT File Allocation Tables

• FAT (File Allocation Table) is a way to track file data on a disk.
• It records where data for different files begins and ends on a hard drive.
• Files can be fragmented, and the FAT organizes those data chunks.

NAS (Network Attached Storage)

• NAS is a storage device that is dedicated to providing data sharing.
• A NAS device acts as a server that handles data processing, but data storage is on a NAS device separate from the physical server.
• NAS devices don't have the same server functions as a server, (no email, authentication, etc) but serve as a central storage point,
• NAS can be expandable (hot upgrades).

Why NAS?

• NAS improves efficiency, provides flexibility, centralizes storage, simplifies management, and has scalable options.
• NAS offers high availability and provides security integration to the server environment, using user authentication and authorization
• NAS devices handle data storage and sharing, and the primary server handles the processing of data and tasks related to operating the server.

Storage Area Networks (SAN)

• SAN is a separate storage network that is independent from the primary network (LAN).
• Data access flows through a dedicated network.
• Provides higher speed and capacity for high-volume access, often used in video broadcasting.
• SANs are frequently used for high-volume data storage, access, and sharing for more complex tasks that utilize more computing power than a NAS device can provide.

SAN Architecture

• SAN uses Fiber Channel for higher network speed to interconnect storage resources.
• The SAN architecture has a separate, dedicated network for storage resources.
• SANs have a fiber channel switch and Disk Array with SAN Controller(s).

Cloud Computing

• Cloud computing is a collection/group of integrated and networked hardware, software, and Internet infrastructure.
• Applications can access data and services over the Internet.
• Cloud computing hides underlying complexity and details of infrastructure
• Examples include Google Docs, Microsoft Online (in the form of SaaS).

SaaS (Software as a Service)

• SaaS is delivering software applications over the internet "on-demand”.
• There are many SaaS providers, including Google, Microsoft, etc.
• Software is hosted and maintained by the SaaS provider for its registered clients.
• Most users will connect to SaaS applications via web browsers or on-demand applications on devices like phone, tablet, or computer.

Advantages of Cloud Computing

• Improved performance, lower software costs

Challenges of Cloud Computing

• Security concerns, data storage locations, dependence on vendor services and market dominance of large companies.

Cloud Storage

• Large Web companies provide storage capacity on a pay-per-use basis
• Data stored remotely is temporarily cached on local devices.
• Backups and data security are handled by the provider.

Task: Network Storage

• Students are asked to think about and identify three questions that would help with the design of a network storage solution before making any conclusions.