CHAPTER 2: ARCHITECTURES
Course Code: CoSc 4038
Distributed
Chapter objectives Systems
Learners should be able to
Differentiate between software architecture and system
architectures
Understand the different software architectural styles
Understand system architecture types
Centralized
Decentralized
2
Distributed
Architectures Systems
Distributed systems are complex.
In order to manage their intrinsic complexity, distributed systems
should be organized properly.
Organization is mostly expressed in terms of its software components.
Different ways to look at organization of distributed systems –two
obvious ones:
Software architecture – logical organization (of software components
and interconnections)
System architecture – physical realization (the instantiation of softwar
e components on real machines)
3
Distributed
Architectural style Systems
It is about software architecture (Not system Arch)
A architectural style is formulated in terms of
Components,
The way that components are connected to each other,
The data exchanged between components, and finally
How these elements are jointly configured into a system.
A component is a modular unit with well-defined interfaces that is replaceable
within its environment.
A connector is a mechanism that mediates communication, coordination, or
cooperation among components.
It allows for the flow of control between components
E.g., facilities for RPC, message passing,
4
or streaming data.
Distributed
Types of Architectural Styles Systems
Common architectural styles of distributed systems
• Layered architectures
• Object-based architectures
• Data-centered architectures
• Event-based architectures
The common to all technique:
Organize your system into logically different components, and distribute
those components over the various machines.
Goal :
Achieving (at a reasonable level) distribution transparency.
5
Distributed
Layered architectural style Systems
It is hierarchical organization
Components are organized in a layered fashion
Components of a layer makes a down-call to components of the layer below
Only in exception, an up-call is made to higher level component
Each layer exposes an interface to be used by above layers
“Multi-level client-server”
Each layer acts as a
Server: service provider to layers “above”
Client :service consumer of layer(s) “below”
Communication protocol-stacks are a typical examples
OSI Reference model
Figure 2-1. The (a) layered architectural style
TCP/IP 6
Distributed
Layered Architecture: Example Systems
7
Distributed
Object-Based Architectures Systems
Components are objects
Objects are easy to be replaced so long as the interface is not touched
It is less structured and hence a relatively loose organization
The calling object might not run on the same machine as the called object
Connectors are RPC and RMI
Notes:
Layered and object-based styles are the most
important styles for distributed systems today
8
Distributed
Data-Centered Architectures Systems
Access and update of data store is the main purpose of the system.
Processes communicate/exchange info primarily by reading and modifying data in
some shared repository (e.g database, distributed file system).
Example, web based Distributed systems are largely data centric
Components:
Data store,
Components, that interact with the store
Connectors:
Queries
9
Distributed
Event–based Architecture Systems
Event based architecture supports publish-subscribe communication
Publisher: components that announce data to be shared
Subscriber: components register their interest for published data.
Decouples sender and receiver (asynchronous communication)
Both parties don’t need to be up at the time of communication
Event can be considered as “a significant change in state”
Components:
Can be an instance of a class or simply a module.
Connectors:
Event buses
10
Distributed
System Architectures Systems
The software components, their interactions, and their place
ment leads to an instance of a software architecture,
also called a system architecture.
System architecture are of three types:
Centralized - most components located on a single machine
Decentralized - most machines have approximately the same functi
onality
Hybrid - some combination
11
Distributed
Centralized Architecture Systems
In the basic client-server model, processes in a distributed system are
divided into two (possibly overlapping) groups.
Server:- is a process implementing a specific service E.g File server
Client:- is a process that requests a service
Clients and servers can be on different machines
Clients follow request/reply model with respect to using services
12
Distributed
Cont... Systems
Communication between a client and a server can be implemented by :
Connectionless protocol when the underlying network is fairly reliable like local-area networks (UDP)
Connection-oriented protocol in WANs, (TCP)
Connectionless communication is efficient
Simply packages a message for the server, identifying the service it wants, along with the nece
ssary input data
But, it is hard for a sender to detect if the message is successfully received
Failure of any sort means no reply
Possibilities:
Request message was lost
Reply message was lost
Server failed either before, during or after13performing the service
Distributed
Cont ... Systems
Common approach to deal with lost request in connectionless communication:
Re-transmission (resending request )
Good for idempotent operations, i.e., operations that could be repeated more than once without ha
rm. E.g., “Return current value of X”
Not good for non idempotent operations like “ increase value of x by 100”
Because, may result in performing the operation twice
In this case reporting an error is appropriate, than resending
For these reason many distributed systems use connection-oriented protocols
Not good enough in LAN as it is slow
However, it fits the unreliable WAN environment
Example, Virtually all internet applications are based on TCP/IP connections
14
Distributed
Application Layering Systems
The client-server model has been subject to many debates and controversies
One issue was how to draw a clear distinction between a client and a server
Many client-server system can be divided into three levels
The user-interface level
The processing level
The data level
These is also called MVC
(Model, View and Controller)
architecture.
Model:- handles all the data-related logic
View:- implements the UI
Controller: intermediary between the model and the view
15
Distributed
Logical Architecture vs. Physical Architecture
Systems
Layer and tier are roughly equivalent terms, but
Layer typically implies software and
Tier is more likely to refer to hardware.
Logical organization is not physical organization.
Physical architecture may or may not match the logical architecture.
Meaning, logically separate components might reside on single machine or on different machines
Clients and servers could be placed on the same node, or be distributed according to several differe
nt topologies.
Single-Tier Architecture: dumb terminal/mainframe configuration
Two-Tier Architecture: client/single server configuration
Three-Tier Architecture: each layer on separate machine
Two-tier and three-tier are the most common
16
Distributed
Two-Tiered Architecture Systems
Where are the three application-layers placed? ( On the client, or on the server machines? )
A range of possible solutions:
Thin-Client- A client machine only implements (part of) the user-interface level
A server machine implementing the rest, i.e, the processing and data levels
Pros: easier to manage, more reliable, client machines don’t need to be so large and powerful
Con: perceived performance loss at client
Fat-Client - All user interface, application processing and some data resides at the client
Pros: reduces work load at server;
More scalable
Cons: harder to manage by system admin,
Less secure
Other solutions in between thin-client and fat-client
17
Distributed
Two-tiered Architectures Systems
Thin client -------- Fat-client
18
Distributed
Three-tiered Systems
The server tier in two-tiered architecture becomes more and more distributed
A single server is no longer adequate for modern information systems
This leads to three-tiered architecture
Server may acting as a client
Three-tiered: each of the three layers corresponds to three separate machines.
19
Distributed
2. Decentralized Architectures Systems
Placing logically different components on different machines is called
vertical distribution
User-interface, Processing components and a data level are on different machin
e
It is similar with the concept of vertical fragmentation in distributed database w
here
Tables are split into column wise and distributed on different
machines
The advantage of VD is that each machine can be tailored for specific type of
function
20
Distributed
Cont… Systems
An alternative to VD is horizontal distribution
A client or server may be physically split up into logically equivalent parts
Each part operates on its own share of the complete data set,
This results in balanced work load
Again this one is similar with that of horizontal fragmentation in distributed database wher
e
Tables are split row wise, and subset of rows distributed onto different mac
hines
Peer-to-peer systems are a class of modern architectures that support h
orizontal distribution.
The functions that need to be carried out are represented by every process that constitute the distrib
uted system
21
Distributed
Peer-to-peer systems Systems
P2P systems partitions tasks or work loads between peers
Often, the processes that constitute the system are all equal
Nodes act as both client and server;
Much of the interaction is symmetric.
Advantages of peer-to-peer system
Have better load balancing
More resistant to denial-of-service attacks,
But, harder to manage than client-server systems.
22
Distributed
Overlay network Systems
Nodes of the P2P distributed system are connected using overlay network
It is network that is built on top of another network
Nodes are formed by the processes of the network.
Overlay networks in the P2P system:
Define the structure between nodes in the system.
Allow nodes to route requests to locations that may not be known at time of request.
The main question for peer-to-peer system is
How to organize the processes in an overlay network
Their organization can be:
Unstructured P2P:
Structured P2P:
Hybrid P2P:
23
Distributed
Unstructured P2P architecture
Systems
Largely relying on randomized algorithm to construct the overla
y network
Each node has a list of neighbours, which is more or less constr
ucted in a random way
One challenge is how to efficiently locate a needed data item
The two common approaches are
Flooding
Random walk
24
Distributed
Cont… Systems
Flooding:
Issuing node u passes request for data d to all neighbors v.
Request is ignored when receiving node had seen it before.
Otherwise, v searches locally for d (recursively).
Return d if found,
Otherwise forward the request to the neighbors
However, this approach causes high signalling traffic over the networ
k
May be limited by a Time-To-Live: a maximum number of hops.
25
Distributed
Cont… Systems
Random walk:
Issuing node u passes request for d to randomly chosen neighbor, v.
If v does not have d, it forwards request to one of its randomly chosen
neighbors, and so on.
26
Distributed
Structured P2P Systems
Nodes are organized following a specific distributed data structure.
The most common one is distributed hash table (DHT)
In such systems, each data item is uniquely associated with a key, in turn used as an
index.
Each node is responsible to store data that are associated with subset of these key
P2P system now responsible for storing (key, value) pairs
Looking up data d with key k means routing request to node with identifier k.
Example
chord
27
Distributed
3. Hybrid Architectures Systems
Many distributed systems require properties from both client-server and peer-to-peer architec
tures.
So, they put together features from both centralized and decentralized architectures, resultin
g in hybrid architectures.
Some nodes are appointed special functions in a well organized fashion. Examples:
Edge-server systems: placed at the edge of enterprise network
E.g., ISPs, which act as servers to their clients, but cooperate with other edge servers to host
shared content
Collaborative distributed systems:
E.g., BitTorrent, which supports parallel downloading and uploading of chunks of a file.
First, interact with client-server system to download the torrent file, and then operate in dece
ntralized manner.
28
End of Chapter-2
29
