Critical Infrastructure Security: The Emerging

Smart Grid
Cpt S 580-04, Cpt S 483-02, EE 582-03, EE 483-01

Dr. Anurag K. Srivastava, Dr. Carl Hauser, Dr. Dave Bakken

Computation Lecture 4: Overview of Distributed Computing (1/2)

Todays Content
1. Administrivia
2. Intro to Distributed Computing

3. Example Local vs. Remote Object Invocation

4. Examples (1.2)
5. Trends (1.3)

6. Focus on Resource Sharing (1.4)

Administrivia
Computation segment HW coming in the next week

Introduction
A distributed system is one in which hardware
or software components located at networked
computers communicate and coordinate their
actions only by message passing

Very broad definition

Lots of examples
Lots of kinds

Abbreviations

Distributed System by DS,

Distributed Computing is DC

You know you have one when the crash of a

computer youve never heard of stops you from
getting any work done. Leslie Lamport

Advantages of Distributed Systems

Share resources (key)
Share devices
Better hardware cost/performance than supercomputers,
multiprocessors
Allows access from many remote users using their simple
PCs (and mobile devices)
Allows for incremental growth (if done right)
Increases reliability and availability (if done right)
Some applications and services are inherently distributed
Can spread the load of a given service much more easily
Can potentially increase security (!!!???)

Consequences of Distributed Systems

Concurrency

Concurrent use of low-level resources: processing, storage

(memory+disk), communications
Mutual exclusion and other synchronization required
Access to resources for a given user often best-effort

No global clock

Cannot often know the exact ordering of events: which happened

first

Independent failures

No longer all or none failures for your program!

Some computers still running, while others failed or partitioned
Failure of a component you are using may not be a clean failure

Variability

Performance of resources (network, CPU, storage) may vary widely

across a remote call chain

Todays Content
1. Administrivia
2. Intro to Distributed Computing

3. Example Local vs. Remote Object Invocation

4. Examples (1.2)
5. Trends (1.3)

6. Focus on Resource Sharing (1.4)

Example Local Call

Caller:
// declare and init stuff
x = new int [100];
y = new util; // create and rtn ref
flag = y.sort(x, 100);

Callee:
// declare and init stuff (not shown)
int util:sort(int [] a, int max) {
// implementation of sort... bubble sort, quicksort,
return status;
}

Example Local Call (BACKUP SLIDE)

Caller:
// declare and init stuff
x = new int [100];
y = new util;
flag = y.sort(x, 100);

Callee:
// declare and init stuff
int util:sort(int [] a, int max) {
// implementation of sort...
return status;
}

Potential assumptions:

Object Invocation conventions between caller (client) and callee

In same address space (on same computer)
In same programming language (usually)
Written by same programmer (often, not always)
Same operating system for both caller and callee
Same CPU type for both caller and callee
Can transfer data and control quickly, effectively in zero time
Both fail, or neither do (for the most part)

None of these assumptions are always true in a

distributed system!

Reminder: Assembler
Example C-like call

X = 4 + ((Y * 4) / (A +
B));

Equivalent assembler (vars on

stack)

ldr r1, [sp, Y] !load Y

mul r1, r1, #4 !Y * 4
ldr r2, [sp, A] !load A
ldr r3, [sp, B] !load B
add r2, r2, r3 !A + B
div r1, r1, r2 !divide the two
add r1, r1, #4 !add four to
result
str r1, [sp, X] !store result in
! X on stack

Reminder: Calling Conventions

To call a function or routine you need to push arguments on

the stack (in the right order), push the return address on the
stack, .., branch to the routine, .
Calling conventions define this for a given compiler/language
High-level language compilers do all this for you
Have to program yourself if using assembler
Calling myFunc() in C:
int main() {
int x = 1;
int y = 2;
int z = myFunc(x,y);
}

int myFunc(int x, int y) {

return x + y
}

myFunc:Reminder: Calling
movl %edi, -4(%rbp)
movl %esi, -8(%rbp)
add
%esi, %edi
movl %esi, %eax
ret
.globl main
main:
movl $1, -4(%rbp)
movl $2, -8(%rbp)
call
ret

myFunc

Conventions

!grab x off stack

!grab y off stack
!add x and y
!return x + y

!x = 1
!y = 2

Example Local Call (2)

Potential assumptions between caller and callee:
Assembler

calling conventions
In same address space (on same computer)
In same programming language (usually)
Same operating system
Same CPU type
Can transfer data and control quickly, effectively in zero
time
Both fail, or neither do (for the most part)

None of these assumptions are always true in a

distributed system!

Example Remote Call

Caller:
// declare and init stuff
x = new int [100];
Y = new util.lookup();
Flag = y.sort(x, 100);

// proxy or stub
// generated by middleware
int util:sort(int [] a, int max){
// put a[], max into struct
// send message with struct
// wait: message w/ struct
// copy from struct to a[],
// status
return status;
}

Callee:
// declare and init stuff
int util_impl:sort(int[] a, int max){
// implementation of sort
return status;
}

// skeleton generated
// by middleware compiler

// receive message with struct

// copy from struct to a[], max
flag = z.sort(a, max)
// copy a[], flag into struct
// send message with struct

Many Local Call Assumptions dont

Hold!
Not a local object Invocation, so need more help
Need

remote equivalent of local (assembler) calling

conventions
In this class we will come to understand this plumbing
much better

Not in same programming language (cant assume)

Not written by same programmer
Not running same operating system for caller and
callee
Not same CPU type for caller and callee

Many Local Call Assumptions dont

Hold! (2)
Not always in the same administrative domain

Latency for transfer of control and data can be large

and, worse, unpredictable
Partial failures

Membership of the system (the computers in its

collection) can change
Unreliable or insecure communication

Bottom Line on Distributed Systems

I dont think we
are in Kansas
anymore, Toto!
Goal of these two lectures is to gain
a basic understanding of:
How and why you are no longer in
Kansas
Where to dig more to find info
about what you can do about it!

Todays Content
1. Administrivia
2. Intro to Distributed Computing

3. Example Local vs. Remote Object Invocation

4. Examples (1.2)
5. Trends (1.3)

6. Focus on Resource Sharing (1.4)

Figure 1.1 Selected app domains and

associated networked apps
Finance and commerce

eCommerce e.g. Amazon and eBay, PayPal,

online banking and trading

The information society

Web information and search engines, ebooks,

Wikipedia; social networking: Facebook and
MySpace.

Creative industries and

entertainment

online gaming, music and film in the home,

user-generated content, e.g. YouTube, Flickr

Healthcare

health informatics, on online patient records,

monitoring patients

Education

e-learning, virtual learning environments;

distance learning

Transport and logistics

GPS in route finding systems, map services:

Google Maps, Google Earth

Science

The Grid as an enabling technology for

collaboration between scientists

Environmental management

sensor technology to monitor earthquakes,

floods
or Distributed
tsunamis
Instructors Guide for Coulouris, Dollimore, Kindberg
and Blair,
Systems: Concepts and Design Edn. 5
Pearson Education 2012

E.g., Google Infrastructure

Underlying physical infrastructure: huge number of
networked computers at data centers all over the
world
Distributed file system
Designed

for very large files

Heavily optimized for style of usage by google apps
(high rates)

Associated structured distributed storage system:

fast access to very large datasets
Lock service: distributed locks and agreement
Programming model supporting managing very large
parallel distributed computations across
infrastructure

Massive Multiplayer Online Games (MMOGs)

Widely-spread users interacting via a persistent
virtual world
Huge challenge to engineer
Fast

response times for a users actions

Scalability
Propagating updates to other users
Consistency

Number of solution kinds

Client-server:

single state in one location (clever server)

Distributing state to many servers (allocate users by
usage patterns)
Research into more radical: peer-to-peer

Financial Trading
HUGE amount of money traded automatically
Ergo

on the cutting edge of distributed systems

technology

E.g., share prices, trends, economic and

political development,
Subscribe to items of interest, an event
Publish-subscribe
Distributed

event-based systems

Fig. 1.2 An example financial trading system

CEP: compose event occurrences into logical, temporal,

or spatial patterns
Trading strategies NOT hard coded into the architecture!

Looking for patterns that indicate a trading opportunity

Instructors Guide for Coulouris, Dollimore, Kindberg and Blair, Distributed Systems: Concepts and Design Edn. 5
Pearson Education 2012

Trading strategies example

WHEN

MSFT price moves outside 2% of MSFT Moving Avg

FOLLOWED-BY (
MyBasket moves up by 0.5%

AND
HPQs price moves up by 5%
OR
MSFTs price moves down by 2%
)
)
ALL WITHIN

any 2 minute time period

THEN
BUY MSFT; SELL HPQ

Todays Content
1. Administrivia
2. Intro to Distributed Computing

3. Example Local vs. Remote Object Invocation

4. Examples (1.2)
5. Trends (1.3)

6. Focus on Resource Sharing (1.4)

Mobile and ubiquitous computing

Computing devices smaller and smaller
Laptops
Handhelds
Wearable

& implantable devices

Embedded in appliances

Mobile computing: performing tasks when not at

home
Includes

location-aware or context-aware

Ubiquitous computing: harnessing lotsa small and

cheap devices in users environment
Useful

only when they can communicate!

Some overlap, but distinct areas mostly

Figure 1.4 Portable and handheld

devices in a distributed system

With right toys, can get lotsa work done away from
home: work, hotel room, partner company,
Support spontaneous interaction! Service discovery
Instructors Guide for Coulouris, Dollimore, Kindberg and Blair, Distributed Systems: Concepts and Design Edn. 5
Pearson Education 2012

Distributed multimedia systems

DEFN: Support range of media types in integrated
manner
Always includes a temporal dimension & preserving
relationships (continuous media)

Access to huge amount of rich info (compared to

text/html)
Places huge demands on distributed infrastructure:
Support

extensible range of encoding and encryption

formats
Provide range of mechanisms to meet users quality of
service (QoS)
Provide appropriate resource management strategies
incl. scheduling
Provide adaptation strategies when QoS cannot be met

Distributed computing as a utility

Grid Computing (e.g., Globus) Cloud computing
Tries to tread distributed resources as a commodity
or utility
Analogies from power grid, networking cloud
Applies to
Physical

resources: storage, processing . Often

virtualized
Software services: email, calendars,
(Seehttp://blogs.computerworld.com/18768/cloud_computing_philosophy)

Clouds usually implemented on (huge) clusters

Example:

Quincy, Washington & Grand County PUD

Figure 1.5 Cloud computing

Instructors Guide for Coulouris, Dollimore, Kindberg and Blair, Distributed Systems: Concepts and Design Edn. 5
Pearson Education 2012

Todays Content
1. Administrivia
2. Intro to Distributed Computing

3. Example Local vs. Remote Object Invocation

4. Examples (1.2)
5. Trends (1.3)

6. Focus on Resource Sharing (1.4)

Focus in resource sharing

Users (and app programmers) want to share higherlevel resources, not lower-level ones
Computer-supported cooperative work (CSCW) AKA
workflow: docs

Service: distinct part of a computer system that

Manages a collection of related resources
Presents their functionality to users & apps (restricted access)

Server: a running program or computer on net that:

Accepts requests from programs on other computers
Performs a service
Responds accordingly

Client (object), server (obj), client-server, remote

invocation
Calls can be nested!

Todays Content
1. Administrivia
2. Intro to Distributed Computing

3. Example Local vs. Remote Object Invocation

4. Examples (1.2)
5. Trends (1.3)

6. Focus on Resource Sharing (1.4)

7. Bonus (if time): Parallel vs. Distributed computing

Comparison: DC and Parallel Computing

(Note: material from: Claudia Leopold, Parallel and
Distributed Computing: A Survey of Models, Paradigms,
and Approaches, John Wiley and Sons, 2001)
Common characteristics

Multiple processors are used

Processors interconnected by some network
Multiple computational activities (processes) are in progress at the
same time and cooperate with each other

Some consider parallel computing a subfield of DC!

Very different.. (e.g., 1995 Kuwait PDC panel)

Parallel computing splits an application up into tasks that

are executed at the same time, whereas distributed
computing splits an application up into tasks that are
executed at different locations using different resources.

Differences: DC and Parallel Computing

Parallel Computing puts emphasis on the
following:
An

application is split into subtasks that are solved

simultaneously, often in a tightly coupled manner
One application is considered at a time, with the goal of
speeding up the processing of that single application
Programs are generally run on homogeneous
architectures, which typically have shared memory
Fault tolerance and security are not generally considered
Or perhaps a checkpoint every hour for a day-long
computation

Differences: DC and Parallel Computing (cont.)

Distributed Computing puts emphasis on the
following:
Computation

uses multiple resources physically

separated: processors, memory, disk, databases
Multiple applications run at a time for many users
Heterogenous systems, open and dynamic
No shared memory, at least not in hardware
Fault tolerance and security must be dealt with (in some
manner)
Sometimes the emphasis is on hiding system internals in
a manner that the distributed system looks like a single
large machine. Feature called a single system image,
used in cluster computing.

Convergence of DC and Parallel

Computing (maybe, eventually)

Architectures (vaguely) approaching each other

Fast

network technologies allow cluster computing

Parallel machines increasingly used as servers in a DS

Parallelism and distribution are closely related

Main

differences in distribution: variable delays and

partial failures

Some joint meetings of parallel and distributed

researchers