Transcript Analyzing Technical Goals and Constraints

CIS 460 - Network Analysis and
Design
Chapter 2
Analyzing Technical Goals and
Constraints
Analyzing Technical Goals
and Constraints
• Goals include:
–
–
–
–
–
–
–
–
–
scalability
availability
performance
security
manageability
usability
adaptability
affordability
Tradeoffs associated with these goals
• Provides terminology to discuss technical goals
with customer
Scalability
• How much growth a network design must support
– maybe a prime goal for some
– Proposed design should be able to adapt to increase in
usage/scope
• Planning for Expansion
–
–
–
–
How many sites to add
How extensive the networks for each new site
How many more users to access the corporate network
How many more servers or hosts to add to the
internetwork
Scalability (Cont’d)
• Expanding the Data Available to Users
– Empowered employees make strategic
decisions that require access to sales,
marketing, engineering and financial data
– The 80/20 rule is no longer valid (80 % stays in
local LANs, 20% destine for other places)
– Increasing access to WWW servers
– Increasing intranet usage
– Strategic alliances with suppliers/customers
Scalability (Cont’d)
• Technical goals for scaling/upgrading enterprise
networks:
– Connect separate department LANs in corporate
internetworks
– Solve LAN/WAN bottleneck problems
– Provide centralized servers that reside on server farms
– Merge an independent SNA network w/enterprise IP
network
– Add new sites to support field offices & telecommuters
– Add new sites to support communication with
customers, suppliers, resellers, and other business
partners
• Constraints
– There are impediments to scalability due to incorrect
Availability
• Refers to the amount of time a network is
available to users and is often a critical goal
• Can be expressed as a percent uptime per year,
month, week, day, or hour
• Also lined to reliability but has a more specific
meaning (percent uptime) than reliability.
Reliability refers to a variety of issues, including
accuracy, error rates, stability and TBF.
• Also associated with resiliency - how much stress
a network can handle and how quickly a network
can rebound from problem.
• Disaster recovery. Disaster recovery plan
Availability (Cont’d)
• Specifying Availability Requirements
– Specify with precision (% uptime), timeframe(day or night
downtime), time unit (day, week, etc.)
• Cost of Downtime
– For critical applications document how much lost per hour
– Also helps to determine whether in-service upgrades must be
supported
• MTBF/MTR
–
–
–
–
Define as mean time between failures and mean time to repair
They can be used to calculate availability goals
4000 hours MTBF is a common goal
Good idea to identify for specific applications in addition to
network as a whole
– Can usually use data supplied by manufacturer for equipment
Network Performance
• Criteria for accepting network performance
– throughput, accuracy, efficiency, delay, and response
time
• Tightly tied to analyzing the existing network to
determine what changes need to be made to meet
performance goals.
• Also tightly linked to scalability goals
• Definitions
– capacity (bandwidth), utilization, optimum utilization,
throughput, offered load, accuracy, efficiency, delay
(latency), delay variation, response time
Network Performance
(Cont’d)
• Optimum Network Utilization
– measure of how much bandwidth is used during a
specific time period. Commonly specified as a
percentage of capacity
– Various tools 7used to measure usage and averaging the
usage over elapsed time.
– Customer may have a network design goal for the
maximum average network utilization allowed on
shared segments.
– For shared Ethernet should not exceed 37 percent else
collision rate becomes excessive
– Toke Ring/FDDI typical goal is 70 percent
– For WANs optimum is also about 70 percent
Network Performance
(Cont’d)
• Throughput
– the quantity of error -free data that is transmitted per
unit of time
– Ideally should be the same as capacity, however nout
usually
– Capacity depends on the physical-layer technologies in
use
– Depends on the access method
• Throughput of Internetworking Devices
– Some specify goals as Number of packets per second
– The maximum rate the device can forward packets
without dropping
Network Performance
(Cont’d)
• Application layer throughput
– a measure of good and relevant application-layer data
transmitted per unit of time (also called goodput)
– Can increase throughput not goodput because extra data
transmitted is overhead data
– Usually measure in kilobytes or megabytes per second
– Constraints
• end-to-end error rates
• protocol functions such as handshaking, windows, &
acknowledgments protocol parameters such as frame size
• protocol parameters such as frame size and retransmission
timers
• PPS or CPS rate of internetworking devices
• Lost packets or cells at internetworking devices
• Workstation and server performance factors
Network Performance
(Cont’d)
• Accuracy
– goal is that data received at the destination must be the
same as the data sent by the source. Typical causes
include power surges or spikes, failing devices, noise,
impedance mismatch, poor physical connection
– Accuracy goals can be specified as a bit error rate
(BER)
– On shared Ethernet errors are often result of collisions
– Collision that happens beyond the first 64 bits of a
frame is a late collision which are illegal and should
never happen
– In token rings accuracy goals sometimes include goals
for minimizing media-access control error reports
Network Performance
(Cont’d)
• Efficiency
– measurement of how effective an 0operation is in comparison to
the cost in effort, energy, time, or money
– Provides a useful way to talk about network performance
– Large frame headers are an obvious cause for inefficiency
– Goal is to minimize the amount of bandwidth used by headers
• Delay and Delay Variation
– Users of interactive applications expect minimal delay in
receiving feedback
– Multimedia applications require a minimal variation in the
amount of delay that packets experience
– Telnet protocol applications are also sensitive to delay because
of echo feedback
Network Performance
(Cont’d)
• Causes of Delay
– delay is relevant to all data transmission technologies, but
particularly satellite links and long terrest4rial cables. Also
time required to put digital data on a transmission line. Packet
switching delay, router delay, queuing delay
• Delay Variation
– Digital voice and video applications are effected with jitter, or
disrupted communications. Can be minimized using a buffer to
minimize jitter
• Response Time
– the network performance goal that users care about most. They
recognize small changes in the expected response time. If less
than 100 ms most users do not notice
Security
• Overall goal is that security problems should not
disrupt the company’s ability to conduct business.
• Have protection so that business data and other
resources are not lost or damaged
• First task is planning which involves analyzing
risks and developing requirements.
– Hackers, financial costs, sensitive of customer data
Security (Cont’d)
• Security Requirements
– protect resources from being incapacitated, stolen,
altered, or harmed.
– Let outsiders access data on public web but not internal
data
– authorize/authenticate branch-office users, mobile users
& telecommuters
– detect intruders
– physically secure hosts and internetworking devices
– protect applications and data from viruses
– train network users and network managers on security
risks
Manageability
• Includes the following network management
functions
• performance management
• fault management
• configuration management
• security management
• accounting management
Usability
• Refers to the ease-of-use with which
network users can access the network and
services
• some network design components can have
a negative effect on usability
Adaptability
• A good network can adapt to new
technologies and changes
• New protocols, new business practices, new
fiscal goals, new legislation or other
possibilities
• Effects it availability
• A flexible design as enables a network to
adapt to changing traffic patterns
Affordability
• Sometimes called cost-effectiveness
• Primary goal is to carry the maximum amount of
traffic for a given financial cost
• In a campus network low cost of often the primary
goal
• For enterprise networks availability is usually
more important than low cost
• Minimizing hiring, training, and maintaining
personnel to operate and manage the network is
also an important goal
Affordability (Cont’d)
• Making Network Design Tradeoffs
– Need to know how important affordability is to other
goals
– Availability often requires Redundant components
which raises costs
– To implement affordability might mean availability
must suffer
– Your customer should identify a single driving network
design goal
• Sometimes making tradeoffs is more complex than
what has been described because goals can differ
for various parts of an internetwork
Technical Goals Checklist
• You should use a technical goals checklist
and Table 2-3 to determine if all the client’s
technical objectives and concerns are
addressed.