Download Subject Review - The CLOUDS Lab

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Parallel port wikipedia , lookup

Distributed operating system wikipedia , lookup

Computer cluster wikipedia , lookup

Transcript
678 Topics Covered (1)

Part A: Foundation




Part B: Cluster Computing





1
Socket Programming
Thread Programming
Elements of Parallel Computing
Elements of Cluster Computing
Cluster Architecture and Components
Single System Image: Concepts and Levels
Parallel Programming Models and Paradigms
MPI Programming
678 Topics Covered (2)

Part C: Grid and Cloud Computing








2
Elements of Grid Computing
Grid Resource Management and Grid Economy
Enterprise Grids and the Aneka Middleware
Nimrod-G Grid Resource Broker
Economic Scheduling Algorithms
Gridbus Broker for Data-Intensive Computing
Globus Toolkit
Cloud Computing and Aneka
Exposure to Industrial Developments

Unique “feature” of this subject:


Industry Speakers (one or more)






Seamless scaling from Clusters to Grids
Story on idea successful commercialisation
Amazon EC2/S3


3
Cluster Computing
Commercial Applications
Sun Grid Engine (SGE)


Microsoft (HPC Specialist)
Sun (Senior Director of Grid Computing)
Amazon (Architect, Amazon Cloud Computing)
Microsoft HPC Server


CLOUDS Lab supported speakers
Computing as utility (pay-as-you-go services for IT)
AWS Marketplace
678 Practical Skills Acquired (1)

Part A: Foundation




Socket Programming
Thread Programming
How to write multi-threaded servers and clients
Part B: Cluster Computing

Parallel Programming:



4
MPI Programming
Coordination of multiple processes and load balancing
Performance Evaluation on real Clusters
678 Practical Skills Acquired (2)

Part C: Grid/Cloud Computing




Report Writing
Team-based projects
Medium scale software development
Developing skills in “Grid” computing technology:





5
Writing a simple Grid resource management system
Client with coordination ability
QoS-based scheduling
User-Interface design for network applications
Cloud Computing (Concepts and Industry Trends)
Quick Revision
678 Topics Covered (1)

Part A: Foundation



7
Socket Programming
Thread Programming
Elements of Parallel Computing
Java Sockets
ServerSocket(1254)

server
Output/write stream
Client
Input/read stream
8
Socket(“128.250.25.158”, 1254)
It can be host_name like “mandroo.cs.mu.oz.au”
Parallel Computing Elements
Applications
Programming paradigms
Threads Interface
Operating System
Microkernel
Multi-Processor Computing System
P
P
P
P Processor
9
P
Thread
P
..
P
Process
Hardware
678 Topics Covered (1)

Part B: Cluster Computing





10
Elements of Cluster Computing
Cluster Architecture and Components
Single System Image: Concepts and Levels
Parallel Programming Models and Paradigms
MPI Programming
Cluster Architecture
Parallel Applications
Parallel Applications
Parallel Applications
Sequential Applications
Sequential Applications
Sequential Applications
Parallel Programming Environment
Cluster Middleware
(Single System Image and Availability Infrastructure)
PC/Workstation
PC/Workstation
PC/Workstation
PC/Workstation
Communications
Communications
Communications
Communications
Software
Software
Software
Software
Network Interface
Hardware
Network Interface
Hardware
Network Interface
Hardware
Network Interface
Hardware
Cluster Interconnection Network/Switch
11
What is Single System Image (SSI)?

SSI is the illusion, created by software or
hardware, that presents a collection of
computing resources as one, more whole
resource.


12
In other words, it the property of a system that hides
the heterogeneous and distributed nature of the
available resources and presents them to users and
applications as a single unified computing resource.
SSI makes the cluster appear like a single
machine to the user, to applications, and to the
network.
SSI Levels

SSI levels of abstractions:
Application and Subsystem Level
Operating System Kernel Level
Hardware Level
13
Levels of Parallelism
PVM/MPI
Threads
Compilers
CPU
14
Task i-l
func1 ( )
{
....
....
}
a ( 0 ) =..
b ( 0 ) =..
+
Task i
func2 ( )
{
....
....
}
a ( 1 )=..
b ( 1 )=..
x
Task i+1
func3 ( )
{
....
....
}
a ( 2 )=..
b ( 2 )=..
Load
Code-Granularity
Code Item
Large grain
(task level)
Program
Medium grain
(control level)
Function (thread)
Fine grain
(data level)
Loop (Compiler)
Very fine grain
(multiple issue)
With hardware
Methodical Design or Stages of Parallel
Programs

Partitioning


Communication


15
Flow of information and coordination among tasks that
are created in the portioning stage.
Agglomeration


Decomposition of computational activities and the
data into small tasks – there exist number of
paradigms – e.g. master worker, pipeline, divide and
conquer, SPMD, and speculation.
Tasks and communication structure created in the
above stages are evaluated for performance and
implementation cost. Tasks may be grouped into larger
tasks to improve communication. Individual
communications can be bundled.
Mapping / Scheduling

Assigning tasks to processors such that job
678 Topics Covered (2)

Part C: Grid and Cloud Computing








16
Elements of Grid Computing
Grid Resource Management and Grid Economy
Enterprise Grids and the Aneka Middleware
Nimrod-G Grid Resource Broker
Economic Scheduling Algorithms
Gridbus Broker for Data-Intensive Computing
Globus Toolkit
Cloud Computing (Concepts and Industry
Trends)
Grid Resources and Scheduling
User Application
Grid Resource Broker
Local Resource Manager
Single CPU
(Time Shared Allocation)
17
Local Resource Manager
SMP
(Time Shared Allocation)
Grid Information Service
Local Resource Manager
2100
2100
2100
2100
2100
2100
2100
2100
Clusters
(Space Shared Allocation)
workload
Gridbus User Console/Portal/Application Interface
App, T, $, Optimization Preference
Gridbus Broker
Gridbus Farming Engine
Schedule Advisor
Trading Manager
Record
Keeper
Grid Dispatcher
Core Middleware
Grid Explorer
TM
$
TS
GE
GIS, NWS
Grid Info Server
RM & TS
$
$
U
Data
Node
Data
Catalog
C
G
Globus enabled node.
18
G
L
Unicore enabled node.
A
Deadline (D) and Budget (B) Constrained
Scheduling Algorithms
Algorithm Execution Execution Compute
Time (D) Cost (B)
Grid
Cost Opt
Limited by D Minimize
Yes
Cost-Time
Opt
Minimize if
possible
Minimize
Yes
Time Opt
Minimize
Limited by B
Yes
Conservative
-Time
Opt
Minimize
Limited by B,
jobs have
guaranteed
minimum
budget
Yes
19
Data Grid
Yes
Yes
Convergence of Paradigms/Communities 
Realisation of Leonard’s “computer utilities” Vision









Web
Data Centres
Utility Computing
Service Computing
Grid Computing
P2P Computing
Cloud Computing
Market-Oriented
Computing
…
Paradigms
20
+
-Ubiquitous
access
-Reliability
-Scalability
-Autonomic
-Dynamic
discovery &
composiability
-QoS
-SLA
-…
-Trillion $ business
- Who will own it?
Attributes/Capabilities
SES (Subject Experience Survey)

21
Q4, subject is well taught - can be
interpreted as: I feel good about what I
learnt, genuine effort by teachers 
5. Strongly agree
:
4. Agree
:
3. Neutral
:
2. Disagree
:
1. Strongly disagree :