Distributed Computing Systems

1
Distributed Computing Systems

• CSCI 6900/4900

2
Review

• Definition characteristics of distributed
  systems
• Distributed system organization
• Design goals
• Resource availability to users
• Transparency
• Openness
• Scalability
• Security and privacy

3
Scalability Techniques

• Hiding communication latencies
• Asynchronous communication
• Reduce amount of data transmitted
• Distribution
• Spreading work across system
• Caching and replication
• Make copies of data and services
• Balance load
• Avoid hot spots

4
Scaling Techniques
1.4

• The difference between letting
• a server or
• a client check forms as they are being filled

5
Scaling Techniques (2)
1.5
An example of dividing the DNS name space into
zones.
6
Hardware Classification

• Based on address space
• Multiprocessor
• Multicomputer
• Based on communication infrastructure
• Bus
• Switched
• Based on uniformity
• Homogenous
• Heterogeneous

7
Hardware Concepts
1.6
Different basic organizations and memories in
distributed computer systems
8
Multiprocessors (1)

• Key Property All CPUs have direct access to
  shared memory
• Coherent memory Reads and writes are sequential

1.7

• A bus-based multiprocessor.

• Scalability
• Ensuring coherency

9
Multiprocessors (2)

• Crossbar switch

Omega switching network
10
Homogeneous Multicomputer Systems

• Also known as System Area Networks (SANs)
• Bus-based (Shared multi-access N/W such as
  Ethernet)
• Switch-based (Massively Parallel Processors,
  Cluster of Workstations)

Grid

• Hypercube

11
Heterogeneous Multicomputer Systems

• Most widely used
• Individual computers can vary widely
• Some node might be multiprocessor machines or
  homogenous multicomputer machines
• Hierarchical systems (multicomputer systems on
  top of existing multicomputer systems).
• Nodes lack global view
• Cannot assume identical or even similar
  performance
• Providing transparency, performance scalability
  are key challenges.

12
Software Concepts

• Functionalities
• Resource managers
• Hiding intricacies and heterogeneity
• Two kinds of operating systems
• Tightly coupled (distributed operating system)
• Loosely coupled (network operating system)
• Middleware
• Providing transparency for loosely coupled systems

13
OS for Distributed Systems
System Description Main Goal
DOS Tightly-coupled operating system for multi-processors and homogeneous multicomputers Hide and manage hardware resources
NOS Loosely-coupled operating system for heterogeneous multicomputers (LAN and WAN) Offer local services to remote clients
Middleware Additional layer atop of NOS implementing general-purpose services Provide distribution transparency
14
Uniprocessor Operating Systems

• Virtualize physical devices (manages devices,
  protects users)
• Two modes (User, Kernel)
• Two kinds (Monolithic, microkernel)

1.11

• Separating applications from operating system
  code through
• a microkernel.

15
Multiprocessor Operating Systems

• Similar in many aspects to uni-processor systems
• Main extension is how shared memory access is
  handled
• Guard against simultaneous access to provide
  consistency
• Two primitives
• Semaphores
• Two atomic primitives (UP and DOWN)
• Monitors

16
Monitors

• Borrowed from programming languages
• Has data and procedures
• One process can execute at any point of time

monitor Counter private int count
0 public int value() return count void
incr () count count 1 void decr()
count count 1
17
Monitors (2)

• Monitors can be used to conditionally block
  processes
• Producers/consumers scenario

monitor Counter private int count 0 int
blocked_procs 0 condition unblocked public
int value () return count void incr ()
if (blocked_procs 0) count
count 1 else signal
(unblocked)
void decr() if (count 0) blocked_procs
blocked_procs 1 wait (unblocked)
blocked_procs blocked_procs 1 else
count count 1