Distributed Information Systems
Motivation
●
To understand the problems that Web services try
to solve it is helpful to understand how distributed
information systems evolved.
●
While technology has changed, problems stayed
the same.
O verview
distributed Information Systems (IS)
●
design and related aspects
●
architectures
●
communication patterns
●
scaling
Design and related Aspects
Layers of an IS
Example
client
browser
<html>
presentation layer
...
</html>
information system
application locic layer
resource management layer
Presentation Layer
●
here is decided HOW data should appear to the
user
●
sometimes referred to as the client (not entirely
true!!!)
Application Logic Layer
●
Data Processing ('The actual Program')
●
here the algorithms are implemented
●
this Layer is often referred to as
– services
– business logic
– business rules
– server
Resource Management Layer
●
deals with and implements different data sources
of IS
●
is the 'data layer' in a restricted interpretation
(Database Management System)
●
can also be an external system, which recursively
uses other ISs
Designs of IS
●
top-down design
●
bottom-up design
top-down design
●
starts with defining functionality desired by the
client ('toplevel goals')
●
implementation of application logic
●
defining the resources needed by applictation logic
top-down [example]
top-down design
1. define access channels and client
client platforms
2. define presentation formats
and protocols for the selected presentation layer
clients and protocols
information system
3. define the functionality
necessary to deliver the contents
and formats needed at the application logic layer
presentation layer
4. define the data sources and data
organization needed to implement resource management layer
the application logic
top-down design
●
usually created to run in homogenous
environments
●
way of distribution has to be specified
●
results in tightly coupled components:
– functionality of each component heavily depends on
functionality of other components
– design is component based, but components are not
standalone
top-down design
top-down architecture
PL - B
PL - A
top-down design PL -C
PL - A PL - B
●
fig 1.3 s.8
PL -C
AL-C AL-B
AL-B
AL-C AL-D
AL-A
AL-A AL-D
RM - 1 RM - 2
RM - 1 RM - 2
advantages & disadvantages
●
advantages:
– design emphasises final goals of the system
– can be optimized for: functional and non-
functional(performance, availability,..) issues
●
disadvantages
– can only be designed from scratch
– legacy systems cannot be integrated
today few ISs are designed purely top-down
bottom-up design
client
bottom-up design
1. define access channels and
client plattforms
2. examine existing resources presentaion layer
fig 1.4
inform ation system
●
and the functionality they
offer
3. wrap existing resources and application logic layer
integrate their functionality into
a constistent interface
4. adapt the output of the
resource management layer
application logic so that it can be
used with the required access
channels and client protocols
bottom-up design
●
out of necessity rather than choice
●
need to integrate legacy systems and/or
applications
●
results in loosely coupled systems
– independent and
– standolone components
●
most distributed IS are result of a bottom-up
design
●
Web services can make those designs more
efficient, cost-effective and simplier to design
bottom-up design
[fig 1.5 p.10] bottom-up architecture
bottom-up design PL - A
PL - B
PL - A PL - B
PL -C PL -C
AL-B AL-C AL-B
AL-C AL-D
AL-A
AL-A AL-D
RM - 1 RM - 2
wrapper wrapper wrapper
legacy legacy
application application
legacy legacy legacy
system system system
Architecture of an Information
System - 4 types:
●
1 – tier
●
2 – tier
●
3 – tier
●
n – tier
1 – tier Architectures
●
were used decades ago..
●
monolithic Information Systems
●
presentation, application logic, and resource
management were merged into a single tier
●
many of these 'old' Systems are still in use!
Design of 1 – tier Architecture
client
presentation layer
[figure 1 – tier]
information system
application locic layer
resource management layer
1 – tier Architecture
advantages:
●
easy to optimize performance
●
no context switching
●
no compatibility issues
●
no client developement, maintainance and
deployment cost
disadvantages:
●
monolithic pieces of code (high maintainance)
●
hard to modify
●
lack of qualified programmers for these systems
2 - tier Architectures
client
presentation layer
●
fig 1.7 p.12
in fo rm a tio n sy ste m
application logic layer
serv er
resource management layer
2 - tier Architectures
●
separation of presentation layer from other 2
layers (app + resource)
●
became popular as 'server/client' systems
●
thin clients/fat clients
●
RPC (Remote Procedure Call)
●
API (Application Program Interface)
●
need for standardization
advantages & disadvantages
advantages
●
portability
●
no need for context switches or calls between
component for key operations
disadvantages
●
limited scalability
●
legacy problems (blown up clients)
3 - tier Architectures
client
presentation
layer
●
fig 1.10 p. 16
in f o r m a tio n sy ste m
application logic layer
middleware
resource management layer
3 - tier Architectures
●
can be achieved by separating RM (resource
management) from application logic layer
●
additional middleware layer between client and
server
– integration logic
– application logic
●
lead to the introduction of clear RM layer
interfaces
●
good at dealing with intgration of different
resources
3 - tier
client
●
fig 1.11 p. 17
presentation layer
middle-
integration logic application ware
logic layer
client client
resource management
wrapper wrapper wrapper layer
1 - tier 2 - tier 3 - tier
advantages & disadvantages
advantages
●
scalability by running each layer on a different
server
●
scalability by distributing AL (application logic
layer) across many nodes
●
additional tier for integration logic
●
flexibility
disadvantages
●
performance loss if distributed over the internet
●
problem when integrating different 3 – tier
systems
n -tier
2 cases of n – tier
●
systems linked with added connectivity through
the internet
●
resource layer is a full fledged 2 - or 3 - tier
system
n -tier
client
Web browser
Web server
in f o r m a tio n s y s te m
●
fig 1.12 p.presentation
20
layer
HTML filter
application logic layer
middleware
resource management layer
n - tier
remote remote remote remote remote remote
client client client client client ... client
●
[fig 1.13 p. 21] INTERNET
FIREWALL
internal
clients
Web server
cluster
middleware
LAN
application
gateways
logic
middleware middleware
application application
database file logic logic
application
server server
additional RM layers
advantages & disadvantages
advantages
●
better scalability
●
higher fault tolerance
●
higher throughput for less cost
disadvantages
●
too much middleware involved
●
redundant functionality
●
difficulty and cost of developement
gains and losses
with growing number of tiers one gains:
●
flexibility
●
functionality
●
possibilities for distribution
but:
●
each tier increases communication costs
●
complexity rises
●
higher complexity of management and tuning
communication in an IS between
distributed layers/tiers
●
synchronous interactions
●
asynchronous interactions
synchronous interactions
(blocking)
●
fig 1.14 p. 23
invoking
execution thread
invoked execution
thread
request
blocking period
response
asynchronous interactions (non
blocking)
●
fig 1.15 p. 25
invoking
execution thread
invoked execution
thread
thread rem ains active
put queue fetch
fetch queue put
scaling multi tier systems
6 steps
●
understand the application environment
●
categorize your workload
●
determine the components most impacted
●
select scaling techniques to apply
●
apply the techniques
●
reevaluate
●
.. and hope its better :)
what do scaling techniques
improve?
incre a se
ca pa city/ improve shift/re duce
sca ling te chnique spe e d e fficie ncy loa d
use fa ste r ma chine x
cre a te ma chine cluste r x
use a spe cia l ma chine x x
se gme nt the work loa d x x
ba tch re que st x
a ggre ga te use r da ta x
ma na ge conne ctions x
ca che da ta & re que sts x x
use faster machine
●
increases the ability to do more work in a unit of
time by processing tasks more rapidly
●
applies to almost all parts of the system (from
edge servers to database server)
create cluster of machines
●
services more client requests. improves response
time through parallelism
●
applies to Web presentation server, Web
application server, directory and security servers
use special machines
●
improves efficiency of a component by using a
special purpose machine, which is optimized for a
specific function
●
applies to egde server, Web presentation server,
directory and security servers, the network and the
Internet firewall
segment the workload
●
splits up workload into managable chunks to
obtain more predictible response times
●
applies to Web representation server, Web
application server, the data server and the network
batch requests
●
reduces number of requests by defining new ones
that combine multiple requests
●
applies to Web presentation server, Web
application server, directory and security servers,
existing business applications and database
aggregate user data
●
allows rapid access to large customer data
controlled by existing system applications by
aggregating distributed customer data into a
costomer information service
●
applies to the Web presentation server, Web
application server and the network
manage connections
●
minimizes number of connections and eliminates
overhead of setting up connections by sharing a
pool of preestablished connections between the
layers
●
applies to Web presentation server, Web aplication
server and the database.
cache
●
improves performance and scalability and
response time by buffering data flows and
reducing consumption of resources
●
applies to the edge server, Web presentation
server, Web application server, network, existing
business applications and the database.
summary
●
layers of an IS
●
designs of distributed IS
●
evolution of architectures and concepts
●
scaling techniques
