1 / 14

Classification of Distributed Systems Properties of Distributed Systems

Classification of Distributed Systems Properties of Distributed Systems. motivation: advantages of distributed systems classification architecture based on interconnection on memory access design based (OS models) design issues of a distributed system transparency heterogeneity autonomy

lchang
Download Presentation

Classification of Distributed Systems Properties of Distributed Systems

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Classification of Distributed SystemsProperties of Distributed Systems • motivation: advantages of distributed systems • classification • architecture based • on interconnection • on memory access • design based (OS models) • design issues of a distributed system • transparency • heterogeneity • autonomy • others

  2. Disk1 P1 P2 P3 M1 M2 M3 Network P4 P5 M4 M5 Printer4 Disk5 Classification of Operating Systems (cont.) • “True” Distributed Operating System • Loosely-coupled hardware • No shared memory, but provides the “feel” of a single memory • Tightly-coupled software • One single OS, or at least the feel of one • Machines are somewhat, but not completely, autonomous

  3. Why Use Distributed Systems?What are the Advantages? • Price / performance • Network of workstations provides more MIPS for less $ than a mainframe does • Higher performance: • n processors potentially give n times the computational power • Resource sharing: • Expensive (scarce) resources need not be replicated for each processor • Scalability: • Modular structure makes it easier to add or replace processors and resources • Reliability: • Replication of processors and resources yields fault tolerance

  4. MIMD parallel and distributed computers loosely tightly coupled coupled Tanenbaum(date?) multiprocessors multicomputers (shared (distributed / memory) private memory) bus switched bus switched Sequent Ultra- Workstations Hypercube on a LAN computer Classification of MIMD Architectures • Tightly coupled parallel processing • Processors share clock and memory, run one OS, communicate frequently • Loosely coupled distributed computing • Each processor has its own memory, runs its own OS (?), communicates infrequently

  5. M1 M2 Mm P1 P2 Pn Classification of Multiprocessors Based on Interconnection Network • Two main types of interconnection: • Bus • Switch (crossbar, multistage switch) • Bus-based interconnection • Simple • Bus is a broadcast medium • Contention for access to bus (does not scale well) • Complicates caches (need snoopy cache)

  6. M1 M2 Mm P1 P2 switch Pn Classification of Multiprocessors Based on Interconnection Network (cont.) • Corssbar switch: • little contention for memory access — multiplememories can be accessed in parallel • Simple routing • Number of crossbar switches grows quadratically

  7. P1 M1 P2 M2 P3 M3 P4 M4 P5 M5 P6 M6 P7 M7 P8 M8 Classification of Multiprocessors Based on Interconnection Network (cont.) • Multistage switch • Reduced number of switches • Increased communication delay(number of hops • Increased contention for memory access • Complex network

  8. Classification of Multicomputers Based on Interconnection Network Two main types of interconnection: • LAN (local area network) – bus based • Switching network • Grid n2 - nodes arranged as an n x n grid • easy to lay-out • Maximum route proportional to n2 • Most messages take multiple hops • Hypercube – a n-degree hypercube (n-cube) consists of 2n nodes (processors) arranged in an n-dimensional cube, where each node is connected to n other nodes • Maximum route proportional to n • Most messages take multiple hops

  9. Classification of Multiprocessors and Multicomputers, Based on Memory Access • UMA — Uniform Memory Access • Main memory is at a central location • NUMA — Non-Uniform Memory Access • Main memory is physically partitioned, with each partition attached to a different processor • Each processor can access its own memory (fast), or the memory of another processor (slow) • NORMA — No Remote Memory Access • Main memory is physically partitioned, with each partition attached to a different processor • A processor can not access the memory of another processor

  10. Distributed System Models • Minicomputer model • Several minicomputers connected to a network, each with several terminals • Workstation model • Many workstations connected to a network • Particularly useful if users can use remote workstations (process migration) • Workstation-server model • Same, plus more some machines run as servers: file server, print server, etc. • Good resource sharing (printers, etc.), cheap workstations (don’t need big disks) • Processor-pool model • Terminals connect to network, pool of processors connect to network

  11. Goals of a Distributed System:Transparency • Access transparency • User is unaware whether a resource is local or remote • Location transparency • User is unaware of physical location of hardware or software resources • location transparency • user mobility • Migration transparency • User is unaware if OS moves processes or resources (e.g., files) move to a different physical locations • Replication transparency • Resource duplication is invisible to users • Concurrency transparency • Resource sharing is invisible to users

  12. Goals of a Distributed System:Support Heterogeneity • Heterogeneity means “consisting of a number of completely different elements” • Computer hardware heterogeneity • Different computer architectures (e.g., instruction sets, data representations) of components in distributed systems • Network heterogeneity • Different transmission media, signaling, network interfaces, and protocols • Software heterogeneity • Different operating systems, application programs

  13. Goals of a Distributed System:Right Degree of Autonomy • Autonomy is a measure of the independence of the components in a distributed system • Low degree of autonomy = dependent • Inflexible • Little robustness in the presence of failures • High degree of autonomy = independent • More flexibility • High redundancy • May still require some central control • Poor resource sharing and coordination

  14. Other design issues of a distributed system • fault tolerance – system should be able to withstand failure of its components and continue (in possibly diminished capacity) its operation • flexibility – to ease modification and enhancement • scalability – system’s performance should not dramatically deteriorate as the system size decreases • security • the communicating parties should be sure of each other identity (be able to trust each other) • the communicating parties should be sure that the communication is not compromised (altered or eavesdropped) • hard because there is no single point of control/authentication

More Related