Hardware and Software Concepts

1
Introduction
Chapter 1

• Part II
• Hardware and Software Concepts

2
Hardware Concepts (1)

• A DS machines connected together
• How are these machines connected together?
• Review multiple-CPU computer Characterization

3
Hardware Concepts (2)
1.6
Different basic organizations and memories in
distributed computer systems
4
Multiprocessors (1)
A bus-based multiprocessor
1.7

• Traffic Problems
• Caching (hit ratio, consistency)
• Scalable?

5
Multiprocessors (2)
1.8

1. A crossbar switch
2. An omega switching network

6
Multicomputers

• Each node is an autonomous machine
• Private memory
• Lower traffic than multiprocessors
• CPU-CPU versus CPU-Memory traffic
• Homogeneous or Heterogeneous

7
Homogeneous Multicomputers (1)

• Similar nodes
• Same processors and memory space
• Homogeneous access to network
• Single network
• Bus-based or point-to-point communication

8
Homogeneous Multicomputers (2)
1-9

1. Grid
2. Hypercube

9
Heterogeneous Multicomputers

• Different nodes
• Nodes can be complex systems
• Non-homogeneous access to network
• Different networks
• DS are commonly built on this hardware category
• Need software to make it transparent

10
Software Concepts

• A DS is similar to an OS
• Resource management
• Virtual machine
• OS
• Tightly coupled
• Loosely coupled

11
Uniprocessor Operating Systems
1.11

• Separating applications from operating system
  code through a microkernel.

12
Multiple CPU Operating 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

• An overview between
• DOS (Distributed Operating Systems)
• NOS (Network Operating Systems)
• Middleware

13
Multiprocessor OS (1)

• Similar to uniprocessor
• Single ready queue
• Mutual exclusion through monitors and semaphores
• Multiple CPUs

14
Multiprocessor DOS (1)
15
Multiprocessor DOS (2)
CPU
CPU
CPU
Kernel
16
Multiprocessor DOS (3)

• A blocks or time-quantum expires

• CPU 1 runs the scheduler and chooses D

A (WR)
A (X)

• Schedule A again on which CPU??

B (X)
C (X)
D (R)
D (X)
C
E (R)
B
A
D
OS
17
Java Monitors (1)
public class Counter private int count 0
public synchronized void reset() count 0
public synchronized int value() return count
public synchronized void inc() count count
1 public synchronized void dec() count
count 1

• A monitor to protect an integer against
  concurrent access.

18
Java Monitors (2)
Lock object
Counter Object
public synchronized void reset() count 0
public synchronized void reset()
acquire(this.lock) count 0 release(this.lo
ck)
19
Multicomputer DOS (1)
1.14

• General structure of a multicomputer operating
  system

20
Multicomputer DOS (2)
1.15

• Alternatives for blocking and buffering in
  message passing.

21
Multicomputer DOS (3)
Synchronization point Send buffer Reliable comm. guaranteed?
Block sender until buffer not full Yes Not necessary
Block sender until message sent No Not necessary
Block sender until message received No Necessary
Block sender until message delivered No Necessary

• Relation between blocking, buffering, and
  reliable communications.

22
Distributed-Shared Memory

• Multicomputers are harder to program
• Multiprocessors are easier to program
• Multicomputers are more efficient and better
  scalable than multiprocessors
• DSM offers a reconciliation

23
Page-Based DSM Example (1)

1. Pages of address space distributed among four
   machines
2. Situation after CPU 1 references page 10
3. Situation if page 10 is read only and replication
   is used

24
Page-Based DSM Example (2)
X
Y
1.18

• False sharing of a page between two independent
  processes.

25
Network OS (1)

• Hardware is not homogeneous
• More primitive than DS
• Not fully transparent

26
Network OS (2)
1-19

• General structure of a network operating system.

27
Network OS (3)

• Services
• rlogin
• File transfer
• This led to the client/server organization

28
Network OS (4)
1-20

• Two clients and a server in a network operating
  system.

29
Network OS (5)
1.21

• Different clients may mount the servers in
  different places.

30
DOS and NOS versus DS (1)

• DOS qualifies as DS?
• Computers are not independent
• Easy to use and transparent
• NOS qualifies as DS?
• No single coherent view
• Scalable and open

31
Positioning Middleware
1-22

• General structure of a distributed system as
  middleware.

32
Middleware

• Local system provides
• Local resource management
• Communication mechanism
• Middleware does not manage the local nodes
• Middleware hides heterogeneity
• Should not call local services directly

33
Middleware Types

• File-based (NOS Transparency)
• For traditional files only
• RPC-based
• Hides communication
• Object-based
• Remote Method Invocation
• Document-based
• The Web

34
Middleware Services

• Communication Facilities
• Naming
• Persistence
• Distributed Transactions
• Security

35
Middleware and Openness
1.23

• In an open middleware-based distributed system,
  the protocols used by each middleware layer
  should be the same, as well as the interfaces
  they offer to applications.

36
Comparison between Systems
Item Distributed OS Distributed OS Network OS Middleware-based DS
Item Multiproc. Multicomp. Network OS Middleware-based DS
Degree of transparency
Same OS on all nodes
Number of copies of OS
Basis for communication
Resource management
Scalability
Openness
Very High
High
Low
High
Yes
No
Yes
No
1
N
N
N
Shared memory
Messages
Files
Model specific
Global, distributed
Per node
Global, Central
Per node
Yes
Varies
Moderately
No
Closed
Closed
Open
Open

• A comparison between multiprocessor operating
  systems, multicomputer operating systems, network
  operating systems, and middleware based
  distributed systems.