Frame Relay Service in Today's Enterprise Network Environment
Written and Produced by techguide.com
Sponsored by:AT&T
Table of Contents
Introduction
This Technology Guide addresses the role that Frame Relay has as one of the
fastest growing and successful network services available today. Its widespread
acceptance as a vehicle for information transfer has surpassed industry
expectations because of its well established technology, virtually universal
availability and cost effectiveness. As a result, Frame Relay service has become
a key element in how businesses will meet the mission critical corporate goals
of the new era.
The Guide explains how Frame Relay service supports the emerging
applications and connectivity requirements of today's Enterprise networks. It
explains the various types of Frame Relay implementations and issues such as the
Committed Information Rate and the need to design the service with a realistic
understanding of the network applications.
Building on the maturity of the Frame Relay technology, the Guide explores
those attributes of a carrier that best meet the needs of the customers in
supplying Frame Relay service, including full customer service and Frame Relay
network management support. An important perspective is that Frame Relay is just
one of a variety of services available from a major carrier. Frame Relay is
positioned as one solution within the family of services of leased line and ATM.
For many customers, hybrid networks which use all three service types will best
meet the needs of the corporation.
Back to the top
Why Frame Relay is Needed in Today's Business Environment
The success of today's business enterprise depends heavily on the computing
and communications infrastructure which supports it. This infrastructure, known
as the Enterprise Network, is considered a strategic asset by businesses, and
has changed dramatically over the years as a result of the evolution in data
processing and telecommunications. Because of the advances in computing and
telecommunications, companies are rapidly adopting new ways of doing business,
with the end result that mission critical applications and operations have moved
onto the network and are increasingly dependent on new networking paradigms.
This new business paradigm is most obvious in several key areas.
The Expanded Role of Networks in Business and New Applications
The entire scope of strategic business concerns, from contemporary business
development, marketing and sales to product development, manufacturing and
distribution, are greatly affected by new networking developments. The reality
of direct consumer interaction through the Internet, the wide scale availability
of information rich graphical applications, all change the way in which business
will be conducted. As processing costs have dropped significantly and processing
power has increased, self-contained desktop applications have emerged to enable
the knowledge worker to operate efficiently and independently. These new
applications generally have some fundamental differences from earlier data
applications:
- Higher Bandwidth Requirements-- The new applications demand much
higher bandwidth because the amount of traffic generated by them is so much
greater.
- Expanded Connectivity Requirements--The nature of Enterprise
Networks today is such that there is a greater need for network connec-tivity
among a much larger end user population. These diverse users and systems need
to communicate with the host and database sites and with each other with
varying degrees of frequency. Where, in the past, the network simply connected
internal, specialized functional areas, the new network interconnects
virtually everyone within the corporate structure. In addition, external
communications with a variety of trading partners, customers, suppliers and
distributors is now commonplace. These various users do not share the same
degree of frequency but do potentially share the same need for access into the
heart of the corporate network. Traditional methods of connecting occasional
users to the network involve the use of dial-up lines of relatively low speed,
which are not satisfactory for meeting today's connec- tivity requirements.
Frame Relay, however, is a natural choice for these since it can provide cost
effective connectivity for even the most infrequent user.
- Diversity of Traffic--Today's networks are highly heterogeneous.
Graphical applications, file transfer, transaction processing, E-mail, etc.
all coexist on the same network and often compete for the same networking
resource. This diversity inevitably demands a mix of ways to handle it.
Prioritization, differing urgencies, and support of bursty traffic are all
requirements for contemporary networks.The traffic characteristics of the new
applications in today's enterprise are different from those of traditional
applications in another way. Whether its moving documents and files between
locations, transferring an image, or accessing a remote server, these new
applications tend to be "bursty" and periodically demand higher
bandwidth. "Burstiness" occurs when a serving facility, which might
ordinarily support a relatively homogeneous and even level of traffic, is
accessed by a transaction, such as an image transfer or a database file, that
requires all, or much, of the available bandwidth for a brief time.
Unfortunately, if the network is not properly designed, the bursty traffic can
severely impact the performance of other traffic on the same facility. This
happens if the traffic volumes are such that there is high likelihood of the
bursty event interfering with the more homogeneous traffic. Under these
circumstances, it is impractical to permanently reserve the entire bandwidth
needed by the bursty application since the occurrences may be infrequent. It
is not cost efficient to have such a high bandwidth facility idle for the
relatively long periods of time that might occur between events. This is a
natural opportunity to use the bursty traffic handling capabilities of Frame
Relay service which can accommodate momentary high bandwidth bursts, as well
as being ubiquitous, reliable and cost effective.
Traditional Networks
Traditional legacy networks do not meet the demands of these new
applications in a number of ways. In the past, except for the circuits in the
corporate backbone which might have been T1 or higher, legacy network circuits
have typically been provisioned at less than 56Kbps. The fixed bandwidth of
these arrangements lacks the efficiency to handle bursty traffic variances and
provides no inherent mechanism for supporting prioritization or occasional
users. The increased number of locations in today's networks has added a layer
of network complexity that is costly and difficult to engineer with legacy
networks. In addition, the distributed host and database arrangements have
created routing and connectivity requirements that have been difficult to cost
effectively support with either point to point or multipoint arrangements.
New Organizational Models
In addition to the new network applications, networking and computing
technologies also make it possible to organize businesses differently.
Groupware and remote networking applications have made "telecommuting"
a reality. It is no longer necessary for workers to come to an office merely to
access, generate or share work products in digital form.
Businesses today have organized in flatter, more geographically dispersed
structures built around specialty areas, market convenience and worker
availability. Some firms operate with most of their employees continually in the
field or at home. These diverse organizational models are possible only because
of the availability of technology to tie them together in ways that enable
efficient and comprehensive exchange of information. Newer services such as
Frame Relay make it practical.
The list of contemporary applications used in these new organizational
models includes document imaging, storage and retrieval, medical and scientific
imaging, distributed on line transaction processing, distributed database/client
server, electronic funds transfer, electronic data interchange, electronic mail
and many more. All of these applications, in one way or another, share a new
profile which includes increased traffic volumes, bigger transaction sizes,
distributed traffic patterns and diverse user populations. This fast changing
application mix is complicated by a concomitant restructuring of the business
organizational model.
Furthermore, these contemporary applications are used by a much wider
segment of the corporate community and are broadly distributed amid many more
users than traditional legacy applications. This leads to a move away from
populations of high volume terminals requiring permanent connectivity to a
broader mix of users, including some with less of a need for constant
connectivity. New networks must be able to support both high volume users and
occasional users within the same coherent architecture.
Back to the top
Typical Frame Relay Applications
Frame Relay best suits applications that have the following characteristics:
- Periodic Traffic--Although Frame Relay Service can be a cost
effective alternative to leased lines supporting the constant traffic of
applications such as SNA, the cost efficiency is most pronounced when the
traffic is variable and/or unpredictable.
- Wide Connectivity--This applies when there are many pre-determined
remote locations to be accessed which have a relatively infrequent need to
communicate. For example, there might be a need to transfer large database
files several times a day to a group of remote branches. Frame Relay service
provides a superb facility for maintaining a list of remote sites and quickly
forwarding traffic.
- Large Transaction Sizes--Because access to Frame Relay Service
usually operates at 56Kbps and higher speeds, it is a suitable vehicle for
graphics, document transmission and large file transfers. It supports LAN-LAN
interconnection and other large transmissions, such as medical/diagnostic
image sharing.
- Bursty Transactions--The best application of Frame Relay is for
heterogeneous networks that support a variety of applications, some of which
are quite large, while others are small. This mix allows the efficiencies of
Frame Relay to be fully utilized.
LAN to LAN Connectivity
LAN - LAN interconnection has become essential in business today. Work group
and team sharing approaches to most corporate activities have necessitated the
interconnection of LANs at increasingly high bandwidths. Typically, at a
minimum, LANs are interconnected at 56 Kbps to 1.5 Mbps. Frame Relay offers an
effective and cost efficient way for LAN interconnection.
The Frame Relay Solution
It is clear that the corporate enterprise has changed dramatically in terms
of its networking needs. Complex, bursty traffic, broad connectivity, new
business structure and the growth of LAN connectivity requires a comprehensive
solution that has, as one of its key components, Frame Relay Service.
Frame Relay service is flexible and can be used for a wide variety of
applications. Some customers use Frame Relay service as a simple replacement of
leased lines. This typically results in reduced facility costs and, in some
cases, even better performance, because Frame Relay can provide high
throughputs. In other cases, Frame Relay service can be used as a multiprotocol
transport vehicle to support the migration to decentralized computing
environments.
Back to the top
Frame Relay--Ready for Prime Time
One of the major benefits of Frame Relay service is that it has reached a
level of wide acceptance and broad deployment that makes it an ideal service to
use in business applications today. It has already gone through a deployment
evolution and is truly "ready for prime time". Customers of Frame
Relay service, as supplied by major carriers such as AT&T, are already well
satisfied with its quality and ease of use.
Vendors and service providers have pledged support for frame relay
development and standards. To this end, the Frame Relay Forum, which includes
major providers such as AT&T, was incorporated in 1991 as a non-profit
organization to promote the implementation of frame relay standards, the
implementation agreements to specify options within the standards and to ensure
interoperability. There are dozens of carriers providing Frame Relay Service and
thousands of companies, world wide, currently using it successfully.
The use of these widely accepted standards, approved by the ITU/TS (formerly
CCITT) and ANSI, and the adoption of these standards by manufacturers of
terminal devices and service providers has enabled Frame Relay service to be the
fastest growing telecommunication service in the Enterprise Network.
Back to the top
How Does Frame Relay Work?
Frame Relay service is based on the data link level protocol defined by ANSI
as T 1.618. It provides for the encapsulation of information (traffic) from
terminal devices connected to the network through routers, bridges and Frame
Relay Access Devices (FRADs).
The Frame Relay frame structure ordinarily contains: Starting Flag (1
octet), Address (2 octets), Information Field (variable length up to 4096
octets), Frame Check Sequence (2 octets) and Ending Flag (1 octet). The 2
address octets (16 bits) contain five unique elements:
- Data Link Connection Identifier (DLCI)--This contains the
identification of the PVC used by the network to find the right path and
destination for the frame.
- Command/Response field--Not used.
- Extended Address--Can be used for extended addressing needs.
- Explicit Congestion Notification fields--Used for flow control.
- Discard Eligibility--Frames marked as being discard eligible can be
discarded in the event of congestion. It is used on frames that exceed the
Committed Information Rate (CIR).
Frame relay is an efficient protocol and it has been streamlined to
eliminate the link by link flow and error control, which have been relegated to
higher levels in the protocol stack. (By comparison, X.25 uses the Link Access
Procedure-Balanced [LAP-B] protocol while Frame Relay uses LAP-D, ITU's Q.922
protocol).
Logical Circuits
Frame Relay standards define three types of logical circuits: permanent,
switched, and multi-cast. The permanent virtual circuit (PVC) is the primary way
in which Frame Relay service is currently provided by carriers. Switched Virtual
Circuits (SVC) are not commonly available from Frame Relay service providers at
this time and multi-cast is provided on a proprietary basis.
Permanent Virtual Circuits (PVCs)
Frame Relay Service does not require a permanent leased line circuit between
the sender and receiver. Instead, a permanent virtual circuit (PVC) is
prearranged by the carrier from the sender to the receiver. This permanent
virtual circuit (PVC) is a predetermined logical path through the carrier
network between the two points. It is invoked when a message is sent. A PVC is
permanently "set up"; however, when there is no message traffic, the
assigned bandwidth can be used by other PVCs occupying the same access facility.
The access facility, i.e., a single physical link from the customer to the
carrier, can support multiple PVCs. For example, a customer might have a 56Kbps
access line to the carrier network over which four different PVCs are arranged
to support four different traffic flows originating from the customer location
to four different destinations.
Committed Information Rate (CIR)
Customers subscribe to PVCs, with specific Committed Information Rates
(CIRs). The CIR is the transport speed which the Frame Relay Network will
maintain between service locations when data is presented. Once established,
there is no "set up" each time the PVC is used. Instead, it uses
pre-arranged routes between switches on trunks within the service provider's
network.
The concept of the committed information rate is at the core of the
successful operation of frame relay service, and unfortunately, is one of the
least understood service parameters. It is a prearranged transmission rate that
assures the customer that the PVC will have, at a minimum, that much bandwidth
available to it for transmission of data.
The CIR assigned to permanent virtual circuits is analogous to the speed of
leased lines. Under ordinary conditions, the service assures the sender that the
data will be delivered at the arranged transmission rate. Frames that are sent
within that parameter are all handled by the service. Because multiple PVCs can
occupy the same physical link to the carrier, any one of the PVCs can use the
full bandwidth of the access link, in excess of the actual CIR of the virtual
circuit. Even if, for example, each of the four PVCs only had a CIR of 9,600
bps, any one of them could use the full 56Kbps of the access facility if the
remaining PVCs were idle.
However, bursty frames offered to the network in excess of the CIR are
marked as Discard Eligible (DE) by the source node (FRAD, router, bridge). The
network will admit frames representing excess data over the CIR if there is
available capacity within the frame relay network. As the network becomes
congested, these Discard Eligible frames may be discarded to ensure that the
network does not become overloaded. A critical aspect of selecting the right
Frame Relay service provider is to understand the basic network capacity
assigned by the carrier to its Frame Relay service, how well the network is
designed to absorb bursty traffic flows and the process by which congestion is
managed. With some carriers, the DE frames have a high probability of arriving
at their destinations because of proper traffic balancing and the allocation of
sufficient excess bandwidth to provide the highest quality level of service.
Overbooking the Access Service
One of the more important characteristics of Frame Relay is that the
customer can overbook the access service. That is, the customer can have a
greater aggregate PVC bandwidth than the capacity of the Frame Relay access port
ordered. In the following sketch, there are 14 PVCs with an aggregate bandwidth
of approximately 650Kbps. The access port is only 384Kbps.
These PVCs are provisioned to provide for the expected bandwidth needs of
the traffic flow when it occurs, but, because of the non-coincidence of traffic,
i.e., not all of the PVCs are active at the same time, the access service will
be sufficient, assuming that the overall utilization is under 50%. There is no "rule
of thumb" for this ratio, but each traffic flow must be considered in terms
of its contribution to the overall utilization of the service.
In this example, the customer device could send data over any one of the
PVCs up to the capacity of the access service; in this case 384Kbps, so long as
the other PVCs were idle. In this example, the physical access circuit is T1
simply because access circuits are usually sold in increments of either 56Kbps
or T1. Fractional T1 increments are not ordinarily available from the customer
premises to the local central office.
Supporting Bursty Traffic within the Carrier's Network
One of the concerns with Frame Relay service is the question of how
overbooked traffic is handled as it travels from the customer access port
through the carrier network. If the traffic on the PVC has been presented to the
carrier at rates in excess of the CIR, the traffic is marked as Discard Eligible
and can be discarded in the event of network congestion. This is a serious issue
and is resolved through several techniques. One technique is for the carrier to
assure sufficient capacity to support anticipated traffic. This requires the
carrier to understand the nature of the traffic it proposes to support and to
supply enough bandwidth to carry it without congestion. Another critical
capability of the carrier is to provide comprehensive congestion management so
that in the event that congestion does occur, the network can quickly respond
and reduce the duration and impact of the congestion. One highly regarded
approach is through the use of PVCs which operate within a closed loop feedback
mechanism to control congestion.
PVCs and Closed Loop Congestion Management
PVCs with closed loop congestion management can provide the end user with
sustained bursting capabilities for the PVC beyond the CIR of that PVC so long
as there is spare capacity in the Frame Relay Service network. Closed loop
congestion management operates through a sophisticated algorithm that monitors
the status of the network and adjusts the available bandwidth of PVCs. This
feature allocates excess bandwidth fairly among the users and provides optimum
protection of network resources by metering access to the network via a closed
loop feedback mechanism.
In some carrier's networks, status indicators which reflect the current
utilization of network resources, are updated at every node crossed by each PVC
in the network. Depending on the particular values of the status indicators, the
algorithm increases, decreases, fast decreases, or does not change the rate of
the PVC. However, the indicators will not cause the available bandwidth of the
PVC to be decreased below the CIR.
An example of the status indicators are:
- Cell Level Utilization (cells per trunk).
- Frame Level Utilization (frames per port).
In some carrier's networks, the cell level utilization is monitored at the
cell buffer in the trunk interface of each node. If there are intermediate nodes
in the PVC path, cell level utilization is also monitored in these nodes.
Frame level utilization is monitored at the frame buffer in the destination
port where the frames are reassembled.
The status indicators are monitored at the destination port for each round
trip network delay. The results of the status indicators are fed back to the
credit rate adjustment mechanism in the source node. Then the bandwidth of the
PVC is adjusted in one of the following ways:
- Rate Up: The bandwidth of the PVC is increased as a percentage of
the CIR. This occurs when network capacity exists to carry the increased
traffic load.
- Rate Down: The bandwidth of the PVC is decreased as a function of
the current value. This occurs when there is a high level of traffic in the
network.
- Fast Decrease: The bandwidth of the PVC is decreased from its
current value by a large percentage. This occurs when there is a very high
level of traffic in the network. It will not drop below the CIR.
The bandwidth of the PVC is adjusted by changing the rate at which credits
are delivered to the PVC's credit buffer. Unused credits stored in the credit
buffer provide the burst capability. By adjusting the rate at which credits are
received, the bandwidth of the PVC is adjusted. The fact that the bandwidth of
the PVCs are increased in proportion to the CIR leads to the fairness property
in some carrier's implementation. The unused network bandwidth is shared between
the PVCs in proportion to their particular CIRs.
Benefits of Closed Loop Congestion Management
In some carrier's implementations, PVCs can offer the user sustained
bursting capabilities beyond the CIR, up to the port speed for an extended
period of time, plus proactive congestion avoidance, highly reliable delivery,
and fairness across all network users. Closed loop congestion management
provides optimum protection of network resources by metering access to the
network via a closed-loop feedback mechanism, avoiding congestion and potential
data drop-outs. The result is highly reliable data delivery and throughput,
fairly allocated to all network end-users.
Frame Relay and SNA
For some situations frame relay service can be utilized to integrate
multi-drop SNA networks and parallel router-based networks for client/server
applications, into a single network. While this integration can also be
accomplished via leased lines, frame relay offers the potential for significant
cost savings, with the added promise of improved performance.
However, one of the biggest challenges is to integrate SNA and LAN traffic
on a single network while obtaining the consistent performance necessary for
efficient SNA sessions. To assist users, routers and FRADs support a number of
features with precisely this objective in mind. Router/FRAD technology also
allows customers to consider replacing their remote Front End Processor (FEP)
concentrators with router technology, although extra care needs to be taken in
this process. In general, a good understanding of (a) how routers support SNA
protocols, (b) how LAN protocols differ from SNA, and (c) how frame relay
differs from leased lines in terms of delay characteristics, is necessary to
ensure consistent performance of SNA applications in the presence of LAN
traffic.
While many proprietary protocols exist for supporting SNA with routers/FRADs
over frame relay, there are two key enabling standards--RFC 1490 and Data Link
Switching (DLSw). These are described briefly below. Frame Relay Service
Carriers usually can provide additional support for customers.
RFC 1490 and SNA
RFC 1490 is a Frame Relay Forum standard for encapsulating multiple
protocols on a frame relay PVC. Furthermore, the Frame Relay Forum document,
FRF.3, describes how SNA can be carried directly over frame relay (in contrast
to encapsulating SNA in TCP/IP).
To explain briefly, SNA requires a connection- oriented, reliable data link
layer protocol. On a leased line network, SDLC provides this function. Since
frame relay does not participate in windowing or retransmissions, Logical Link
Layer Type II (LLC2) (IEEE802.2) protocols are used over frame relay
connections.
RFC 1490 essentially defines two methods of transport of LLC2 frames over
frame relay--the Routed Frame Format and the Bridged Frame Format. The routed
frame format encapsulates LLC2 frames directly in frame relay, whereas the
bridged frame format uses Media Access Control (MAC) addresses in addition. The
bridged frame format is less restrictive and easier to configure, but has more
overhead, compared to the routed frame format. Customers should request
assistance from their frame relay service carrier to implement either method
optimally for their specific applications.
DLSw Over Frame Relay
Data Link Switching (DLSw) is used for routing of SNA traffic, also
involving the use of TCP/IP, but as an external transport after the SDLC
protocol issues have been dealt with. Data Link Switching, originally developed
by IBM, has since been adopted by the members of the APPN (Advanced Peer-to-Peer
Networking) Implementers Workshop. It is also described by the Frame Relay Forum
within RFC 1795.
DLSw is essentially a variety of TCP/IP encapsulation used for performing
source route bridging over a WAN. As such, it can be overlaid onto a Frame Relay
network. DLSw provides virtual point-to-point connections between pairs of Data
Link Layer Media Access Control (MAC) addresses. The DLSw specification deals
with the encapsulation of SNA, APPN and NetBios in TCP/IP, which can then be
transported using Frame Relay paths. One of its key functions is to interact
with the native SNA devices at either end of the path and eliminate the protocol
overhead and polling from occupying the network.
Back to the top
Other Frame Relay Service Considerations
Frame Relay service is just one of the many data transport offerings of
major carriers such as AT&T. Frame Relay, although best suited to the types
of applications described above, does not meet the needs of all traffic types.
For applications such as voice or broadcast video, ATM is probably a better
solution. When there is a need to interconnect major host facilities together
for constant, high traffic volume transfers, T1 or T3 leased lines would be more
ideal.
In general, frame relay service is complementary to other services like ISDN
and ATM. Each has strengths that can support a comprehensive customer solution.
Each offers a specific range of bandwidth and capabilities suitable for
particular applications. For example:
- Asynchronous services offer a simple analog interface over dial up or
leased lines at low bandwidth. X.25, for its part, offers better utilization
of bandwidth through sharing and accommodation of the network for relatively
low intensity traffic.
- ISDN is a circuit switched digital technology, offering bandwidths of
56Kbps to T1 speeds, which allows the user to establish high speed dial up
connections around the world. It is used mostly for voice and data. It does
not pass frames across the network according to addresses contained in the
frame. Instead, just like leased lines, once the connection is established,
transmission occurs without the imposition of additional protocol overhead and
there is no need to either encapsulate or unbundle the traffic flow. Billing
for ISDN services is similar to usage based services such as long distance
toll calls. The customer pays for the basic PRI or BRI access circuit and is
billed for usage as it occurs.
- ATM, which is a cell based, switching and transmission technology similar
to Frame Relay, typically operates at higher bandwidth rates. It offers
interfaces at 1.544 Mbps and 45 Mbps and it is planned to reach 622 Mbps and
beyond. It is intended to operate with the rates that will be available
through SONET technology. ATM also uses Virtual Circuit and CIR concepts and
is a superset of protocol support capabilities. It can carry Frame Relay as
well as various LAN systems, Video, FAX and Voice.
Central Office FRADs
Contemporary frame relay service providers have an additional capability to
support other traffic types by routing them in native mode to the nearest frame
relay central office, where the traffic is encapsulated into a frame relay data
stream through frame relay access devices (FRADs). These CO FRADs provide a
cost-effective way to combine point-to-point, legacy network traffic such as
SDLC and Bisync with router-based traffic through protocol encapsulation. CO
FRADs are installed in the carrier CO and allow the traffic to meet its
performance and availability requirements. Typical applications include host
connectivity from ATMs (Automatic Teller Machines), lottery machines,
point-of-sale terminals, branch offices, remote claims adjusters or travel
agents.
Back to the top
Distinguishing Service Providers
Frame Relay is now available as a mature service. International as well as
domestic providers include Frame Relay service in their portfolio of enterprise
network offerings. These services, however, vary from carrier to carrier and
their quality is only as good as the carrier's network and platform
capabilities. The educated Frame Relay customer will understand and consider
some of the issues important to Frame Relay and its provisioning before
committing to a particular provider.
Customers can understand how carrier's implementations differ by comparing
them and their services against these important characteristics. Customers can
make more informed decisions concerning the value received by examining
carriers' Frame Relay services in light of these measures:
Carrier Architecture
Customers planning to implement Frame Relay service should consider the
infrastructure, capability and architecture of the prospective service provider
to determine if that provider has the potential to truly meet their total system
and performance needs. The carrier's architecture should offer critical depth,
such as the ability to support a wide range of service interfaces. Another key
element of the carrier architecture, is the method used to control network
congestion.
Customer Network Management-- Electronic Servicing
One major distinction in Frame Relay service is the ability for the user to
manage the frame relay network on an ad-hoc and as needed basis. In order for
this to work effectively, the carrier must provide two important components:
regular detailed reporting of traffic by PVC and the means to electronically
effect changes in the PVC and port configurations.
Web Based Interface
As the Internet becomes more pervasive and used for virtually all
inter-company traffic, it makes sense that customer network management systems
become available within the context of a standardized browser interface. The
Internet allows users access to regular reports including:
- Port and PVC utilization.
- Discards, CRC error and congestion notifications.
- Exception reports with definable thresholds allowing rapid identification
of problem areas.
Electronic Ordering enables customers to order new frame relay ports and
PVCs and to upgrade service speeds (CIRs). This provides direct management of
network configurations.
Electronic trouble ticketing provides web based reporting for circuit, port
and PVC troubles and allows timing tracking of problem resolution.
A Clean, Seamless Evolution
The carrier should be prepared to offer compatibility and interoperability
between Frame Relay and other services such as ATM. The best Frame Relay service
providers will integrate their Frame Relay service with other transport options
and additional services into seamless offerings that meet critical business
needs. They will offer Frame Relay as part of a complete range of transport
options for LAN interconnect solutions, in which the transport services (e.g.,
Frame Relay, dedicated private line, fractional T1/E1 and 45 Mbps) are combined
in an integrated solution. The carrier should have a strategy for integrating
Frame Relay with ATM in a hybrid mix of services specifically designed to meet
the needs of the customer. Since the Frame Relay specification prescribes only
the access technology, network providers are free to use frame or cell-based
backbones in their networks. Those that use cell-based backbones are positioned
for expansion in scale and scope, and are likely to be better equipped to grow
in the long term.
An Array of Access Options
Customers' needs vary, and some implementations are optimized using a mix of
switched and dedicated access lines at different rates. Carriers should offer
dedicated access at popular rates (e.g., 56/64 Kbps, fractional T1, 1.5 Mbps),
as well as dial access. ISDN is another increasingly important way to access
Frame Relay service. The Frame Relay service should support a wide range of
interface rates (from a few kilobits per second to megabits per second), and
protection in case of local access channel or CPE failure.
Around the Clock Network Surveillance and Monitoring
Maintaining high availability and quality service requires continuous
network surveillance. This includes continuous remote monitoring of all the
physical elements of the network, immediate response to troubles or questions,
and superior problem isolation and repair capability.
Disaster Recovery
Some Frame Relay Service Carriers provide Disaster Recovery to redirect PVCs
to existing alternate access channels in case of local access channel or CPE
failures. This assures the Frame Relay customer of uninterrupted communications
between an enterprise's computer center and remote LAN and distributed
environments. Other disaster recovery options provide backup PVCs or growable
PVCs to alternate data sites or disaster recovery vendors in the event of a site
disaster. Rapid service restoral is possible by temporarily redirecting the PVCs
or expanding the CIR of an existing PVC to an alternate data site to handle the
networking needs in the event of a disaster.
Real-Time Access to Network Operations
Many customers may not want to be involved in the day-to-day operation of
their networks. Instead they look to carriers to handle the operational details
for them. A carrier who excels at this monitors all aspects of network
performance and offers customer support of the CPE, including network design,
configuration support, remote monitoring, and trouble referral and tracking. In
addition, the carrier offers the user the means to monitor their own private
virtual networks for status and performance. They provide real-time access to
information using SNMP (Simple Network Management Protocol), including graphics
based network management work stations. The carrier also provides customers with
a single point of contact to their network operations center for coordinated
problem resolution.
End-to-End Engineering and Planning
There is more to provisioning Frame Relay service than simply securing the
access connections. Customers may prefer to outsource some or all of the work
associated with implementing a Frame Relay network. If so, the following
additional services that can be provided by carriers are likely to meet the
customer's needs:
- Provision and installation of the cabling to connect the CPE to the
network.
- Design and configuration of required customer premises equipment (CPE) to
ensure optimum use of the network.
- Complete installation of both the network and customer premises equipment,
including routers and bridges.
- Provisioning and installation of the CPE.
Internet Access Services
Carriers should offer Internet services, including:
- Internet connectivity using PVCs.
- Internet directory services.
On-Site Support for Problem Resolution
The duration of failures depends on the carrier's ability to be on-site
anywhere within a few hours. Good on-site service should be structured to
provide the customer with optional levels of support, depending on their
priorities.
Predictive Failure Modeling and Preventative Maintenance
While 100% fail safe service is too costly to be practical, it can be
approached using continuous fault monitoring from multiple network operations
centers (NOCs). These centers use remote monitoring systems and predictive
modeling of service degradation to schedule preventative maintenance and perform
capacity management. Trended analysis reports of performance over time can quite
accurately predict when parameters are nearing service affecting thresholds, and
when failures are likely to occur.
Robust, Scalable Backbone Network
- The carrier should be able to smoothly grow and reliably protect the
customer's service through:
- Fully-protected backbone network facilities and switches.
- Evolution path for higher access speeds as customer demands increase.
- Robust design and engineering practices to ensure sufficient capacity.
- Scalable, easily expanded backbone networks to handle growth in demand.
- Interworking services.
Support Sustained Data Bursts
Frame relay service should successfully and consistently support sustained
bursts of data at levels exceeding the customer's committed information rate.
Carriers differ markedly in their ability to support sustained bursts. There
should be fair, proportional allocation of spare bandwidth to ensure that one
customer's traffic doesn't affect the performance of other traffic.
Seamless Interworking and Integration of Multiple Services
While customers may choose to perform their own access coordination and
network management, they should have the option of allowing the carrier to do
that for them. Carriers should offer:
- Access coordination between Local Exchange Carriers (LECs) and alternate
access providers.
- Full access to extensive network management capabilities.
- Integration with existing and new services, including support for protocol
conversion.
- Seamless integration of all access, carrier network and CPE into a unified,
single network.
Support for Servicing and Network Expansion
- Coordination of the addition of new locations into the enterprise network
and reconfiguring the logical router configurations in response to changes in
the LAN configuration.
- Monitoring of the network continuously to ensure optimum use of resources.
Pricing Flexibility
A properly designed pricing structure offers flexibility to the enterprise
network user. In particular, carriers offer:
- A variety of price plans and separate charges for ports and PVCs.
- Long-term service agreement incentive pricing plans.
- Volume discounts.
Back to the top
Determining the Value of Frame Relay Service
The value of Frame Relay service to a particular customer will vary
depending on the network topology, the traffic patterns, and the geographical
distribution of sites. While it is difficult to apply financial analyses
universally, users in a wide variety of businesses report both increased levels
of service and cost savings over private line solutions after implementing Frame
Relay networks. Recent customer surveys have shown:
- Reported savings on the order of 30% are not uncommon. Service improvements
and cost savings are reported by users in business as diverse as financial
services, health care, insurance, manufacturing, publishing, network services,
retail, wholesale, communications, government, transportation, pharmaceuticals
and shipping.
- Users are receiving the performance and redundancy of meshed networks that
they could not otherwise afford.
- Frame Relay service tempers the problems usually associated with building
and managing complex mesh networks.
- The multi-protocol encapsulation capability of Frame Relay service has
enabled the migration from dedicated SNA networks into multi-protocol,
decentralized environments.
- Frame Relay service offers some significant benefits for the SNA community,
especially link consolidation through the nearly limitless virtual circuits
that can be theoretically provisioned on each access link, greatly reducing
access port requirements.
- The higher speed access (56 Kbps and above) significantly improves
performance over traditional 9.6 Kbps leased line and multi-drop networks so
characteristic of the SNA environment.
Much of the Frame Relay savings comes from the reduction of CPE ports and
circuits both in the access and interoffice portions of the network. With
conventional fixed-bandwidth private line services, there is a need for multiple
port terminations on the CPE and a need for multiple, fixed-bandwidth, dedicated
channels in the access connections. This results in poor utilization of network
resources, and increased cost, since the number of connections goes up
exponentially with the number of sites requiring interconnection. The following
comparison of Frame Relay and leased lines illustrates this point.
Consider the simple four node network below. To establish full mesh
connectivity with leased lines requires six lines and twelve ports (three per
customer location), as shown in the following diagram.
Six (6) leased lines and twelve (12) CPE ports are required for full
connectivity.
The same full mesh connectivity can be achieved using Frame Relay service
with just four access lines and four CPE ports (one per customer location). Each
port and each access line contain three PVCs going to each of the other
locations. Thus, Frame Relay affords the capacity efficiencies of star networks
plus the connectivity of mesh networks.
The savings in port and access links becomes substantial as the number of
interconnected nodes increases. In the four node illustrative example, the
number of links was reduced from six to four, and the CPE ports from twelve to
four. The savings achievable increases substantially as the number of nodes to
be interconnected increases, as shown below.
The practical effect of this relationship of links and ports needed as nodes
increase using the private line solution is that full mesh connectivity quickly
becomes cost prohibitive, and it is not implemented fully as the number of nodes
increases. This is not the case with Frame Relay, so Frame Relay makes full mesh
interconnectivity achievable.
Frame Relay--A Solution for the Future
The corporate enterprise has changed dramatically in terms of its overall
networking needs. Complex, bursty traffic, broad connectivity, new business
structure and the growth of LAN to LAN connections mandates a comprehensive
solution that has, as one of its key components, Frame Relay Service.
Frame Relay service is a flexible and exciting way to support the wide
variety of new business applications. Frame Relay service is used as a
multi-protocol support and is becoming the most important component of
tomorrow's networks.
Back to the top
Customer Profile
AT&T Frame Relay Service Speeds Mass Data Transfer, Reduces Customer's
Critical Design Cycle Times
This customer has 40 design centers worldwide, and an employee force
exceeding 5,500 people. It has leading edge production processes that produce
circuit components just microns across. While they hold a position of leadership
in the field, competition is intense. Thus, maintaining leadership while
focusing on bottom line profitability in a global marketplace requires
innovative ways to bring their products to market. One exciting new idea they
used included an AT&T Frame Relay Service offering which helped them reduce
critical design cycles.
Initially, they implemented a system to off-load design work to
underutilized design centers, interconnecting geographically dispersed LAN's.
However, this private line network operating at 19.2Kbps wasn't fast or
efficient enough to transmit the typical 200 to 300 megabyte files. Files had to
be sent by overnight mail. Compounding the problem, they often sent engineers
from one center to another to best use available time and skills.
Working with AT&T Data Networking Solutions to customize a solution,
they found that this new "alliance" could help move large design files
to where they were needed, at very high speeds. They are now able to off-load
design work to underutilized design centers during peak business periods,
helping to get more design work completed in a shorter time frame.
What's more, access time to the headquarters mainframe has been cut from
three or four seconds to sub-second response. Formerly, it would take 24 hours
to update design software and library modules. Now it takes about an hour.
Investment Protection
LAN-to-WAN systems work best when speed, reliable and flexible networking,
economy, and global support are synchronized. For this customer, AT&T Frame
Relay Service got the job done. They had invaluable access to AT&T network
engineering, design and performance analysis. AT&T Laboratories' technical
support was also readily available.
AT&T Has 21st Century Solutions
With increasing competition, this customer required greater connectivity,
higher performance, and improved economics. They had typical Frame Relay
applications including distributed databases, CAD/CAM/CAE, imaging, graphics,
software and information distribution and groupware.
Taking advantage of highly accurate digital networks and intelligent network
endpoints devices, Frame Relay with its protocol conversion capability allows
the transfer of more data more simply and cost-effectively than on otherwise
underutilized private lines. It can enable handling of unpredictable data
traffic, providing each network endpoint the ability to communicate with
multiple destinations.
They were also interested in AT&T Frame Relay Customer Network
Management Services (CNMS). These services provide measurements and analyses of
data traffic and information to better understand the dynamics of network
communications patterns. And, the CNMS measurements allow them to maximize data
networking efficiency; proactively indicating areas where there is a need to
grow or evolve.
More Than Just Faster Transmissions
In addition to faster transmission times, AT&T's Frame Relay Service
enabled this customer to better optimize their data communications network. They
are now able to look at integrated access, bringing together SDN, 800 service
and frame relay. What's more, they now enjoy a greater degree of disaster
protection. If a fire or other disaster should strike one facility, it's now far
easier and faster to shift large amounts of data to another center via Frame
Relay than it was with their previous data network solution.
Back to the top
Glossary of Terms
Access Method--The method by which networked stations determine when they
can transmit data on a shared transmission medium. Also, the software within an
SNA processor that controls the flow of information through a network.
Access Minutes--The usage of exchange facilities in interstate or foreign
service for the purpose of calculating chargeable usage. On the originating end
of an interstate or foreign call, usage is measured from the time the
originating end user's call is delivered by the Telephone Company to and
acknowledged as received by the customer's facilities connected with the
originating exchange. On the terminating end of an interstate or foreign call,
usage is measured from the time the call is received by the end user in the
terminating exchange. Timing of usage at both originating and terminating ends
of an interstate or foreign call shall terminate when the calling or called
party disconnects, whichever event is recognized first in the originating and
terminating end exchanges, as applicable.
Access Rate--The transmission speed, in bits per second, of the physical
access circuit between the end user and the network.
Access Tandem--A Telephone Company switching system that provides a
concentration and distribution function for originating or terminating traffic
between end offices and customer's premises.
Acknowledgment (ACK)--Portion of any communications protocol responsible for
acknowledging the receipt of a transmission.
Adapter--A board installed in a computer system to provide network
communication capabilities to and from that computer system. Also called a
Network Interface Card (NIC).
Adapter Card--Circuit board or other hardware that provides the physical
interface to the communications network.
Asynchronous Transfer Mode (ATM)--The CCITT standard for cell relay wherein
information for multiple types of services (voice, video, data) is conveyed in
small, fixed-size cells. ATM is a connection oriented technology used in both
LAN and WAN environments.
Asynchronous Transmission--Data transmission one character at a time, with
intervals of varying lengths between transmittals. Start and stop bits at the
beginning and end of each character control the transmission.
B Channel--In ISDN, a full duplex, 64 Kbps channel for sending data.
Backbone--The part of a network used as the primary path for transporting
traffic between network segments.
Backward Explicit Congestion Notification (BECN)--A bit in the frame relay
header. The bit is set by a congested network node in any frame which is
traveling in the reverse direction of the congestion. (In frame relay, a node
can be congested in one direction of frame flow but not in the other.)
Bandwidth--Measure of the information capacity of a transmission channel.
Basic Rate Interface (BRI)--ISDN standards and specifications for provision
of low-speed ISDN services. Supports two "B" channels of 64 Kbps each
and one "D" channel of 16 Kbps on a single wire pair.
Bridge--A device that connects and passes packets between two network
segments. Bridges operate at Layer 2 of the OSI reference model, the data-link
layer, and are insensitive to upper-layer protocols. A bridge will examine all
frames arriving on its ports and will filter, forward or flood a frame depending
on the frame's Layer 2 destination address.
Bridge/Router--A device that can provide the functions of a bridge, router
or both concurrently. Bridge/router can route one or more protocols, such as
TCP/IP and or XNS, and bridge all other traffic.
Cell--For ATM, most vendors have agreed that this information "package"
will be developed consisting of 53 bytes or "octets". Of these, the
first 5 constitute the header; 48 carry the payload.
Cellular Digital Packet Data (CDPD)--A wireless packet data service for
mobile users. Provides high speed, secure wide area IP (Internet Protocol)
network connectivity and operates at 19.2 kilobits per second. To provide
connectivity back to the customer's enterprise networks, CDPD networks interface
with existing wireline networks and services, including frame relay and the
Internet.
Channel Service Unit/Data Service Unit (CSU/DSU)--A digital interface unit
that connects end user equipment to the local digital telephone loop.
Client/Server--A distributed system model of computing that brings computing
power to the desktop, where users ("clients") access resources from
servers.
Committed Information Rate (CIR)--The transport speed the frame relay
network will maintain between service locations when data is presented.
Constant Bit Rate (CBR)--Delay intensive applications such as video and
voice, that must be digitized and represented by a continuous bit stream. CBR
traffic requires guaranteed levels of service and throughput.
Contention--Network access method where devices compete for the right to
access the physical medium.
Customer Premises Equipment (CPE)--Terminating equipment, such as terminals,
phones, routers and modems, supplied by the phone company, installed at customer
sites, and connected to the phone company network.
Data Link Connection Identifier (DLCI)--A value in frame relay that
identifies a logical connection.
Dedicated Line--A transmission circuit installed between two sites of a
private network and "open," or available, at all times.
Delay--Amount of time a call spends waiting to be processed.
Dial Up--A type of communication that is established by a switched-circuit
connection using the telephone network.
Digital Signal 0 (DS-0)--North American Digital Hierarchy signaling standard
for transmission at 64 Kbps.
Digital Signal 1 (DS-1)--North American Digital Hierarchy signaling standard
for transmissions at 2.544 Mbps. Supports 24 simultaneous DS-O signals. Term
often used interchangeably with T-1, although DS-1 signals may be exchanged over
other transmission systems.
Digital Signal 3 (DS-3)--North American Digital Hierarchy signaling standard
for transmission at 44.736 Mbps. Supports 28 simultaneous DS-1 signals.
Discard Eligible--A 1-bit field in a frame relay header that provides a two
level priority indicator. Used to bias discard of frames in the event of
congestion toward lower priority frames. Similar to the CLP bit in ATM.
Encapsulation--A protocol technique in which traffic frames of one protocol
type are transmitted unchanged through the facility of a different protocol
based system. The transmitting protocol adds sufficient information to the
offered frame to send it successfully to its destination, where the added
information is stripped away.
Enterprise Network--A geographically dispersed network under the auspices of
one organization.
File Transfer Protocol (FTP)--An IP application protocol for transferring
files between network nodes.
Firewall--Isolation of LAN segments from each other to protect data
resources and help manage traffic.
Forward Explicit Congestion Notification (FECN)--A bit in the frame relay
header. The bit is set by a congested network node in any frame which is
traveling in the same direction as the congestion. (In frame relay, a node can
be congested in one direction of frame flow but not in the other).
Fractional T-1--A WAN communications service that provides the user with
some portion of a T1 circuit which has been divided into 24 separate 64 Kbps
channels. It is known as fractional E-1 in Europe.
Frame--A logical grouping of information sent as a link-layer unit over a
transmission medium. The terms packet, datagram, segment, and message are also
used to describe logical information groupings at various layers of the OSI
reference model and in various technology circles.
Frame Relay--High-performance interface for packet- switching networks;
considered more efficient than X.25 which it is expected to replace. Frame relay
technology can handle "bursty" communications that have rapidly
changing bandwidth requirements.
Frame Relay Access Device (Assembler/ Disassembler) (FRAD)--Functions like a
PAD.
Frame Relay Forum--A voluntary organization composed of Frame Relay vendors,
manufacturers, service providers, research organizations and users. Similar in
purpose to the ATM Forum.
GUI--Graphic User Interface.
Integrated Services Digital Network (ISDN)-- The recommendation published by
CCITT for private or public digital telephone networks where binary data, such
as graphics and digitized voice and data transmission, pass over the same
digital network that carries most telephone transmissions today.
Internetwork Packet Exchange, Network Protocol (IPX)--LAN protocol developed
by Novell for NetWare.
Latency--The delay between the time a device receives a frame and the frame
is forwarded out of the destination port.
Leased Line--A transmission line reserved by a communications carrier for
the private use of a customer.
Load Balancing--In routing, the ability of the router to distribute traffic
over all its network ports that are the same distance from the destination
address. It increases the use of network segments, which increase the effective
network bandwidth.
Local Area Network (LAN)--A network covering a relatively small geographic
area (usually not larger than a floor or small building). Compared to WANs, LANs
are usually characterized by relatively high data rates.
Logical Link Control Type 2 (LLC2)--A connection oriented mode of operation
within the logical link control sub-layer of the ISO data link protocol layer.
Metropolitan Area Network--A data communication network covering the
geographic area of a city (generally, larger than a LAN but smaller than a WAN).
FDDI can provide a private MAN, while IEEE 802.6 can provide a public MAN.
Network Address--Also called a protocol address. A network layer address
referring to a logical, rather than a physical, network device.
Packet--A logical grouping of information that includes a header and
(usually) user data.
Permanent Virtual Circuit (PVC)--A defined virtual link with fixed
end-points that are set-up by the network manager. A single virtual path may
support multiple PVCs.
Point-to-Point Protocol (PPP)--Successor to SLIP; provides router-to-router
and host-to-network connections over both synchronous and asynchronous circuits.
Primary Rate Interface--ISDN interface to primary access, consisting of a
single 64 Kbps D channel plus 23 or 30 B channels for voice and/or data.
Protocol--A formal description of a set of rules and conventions that govern
how devices on a network exchange information.
Protocol Stack--Related layers of protocol software that function together
to implement a particular communications architecture. Examples include
AppleTalk and DECnet.
Router--An OSI Layer 3 device that can decide which of several paths network
traffic will follow based on some optimality metric. Also called a gateway
(although this definition of gateway is becoming increasingly outdated), routers
forward packets from one network to another, based on network-layer information.
Routing--The process of finding a path to the destination host. Routing is
very complex in large networks because of the many potential intermediate
destinations a packet might traverse before reaching its destination host.
Routing Table--A table stored in a router or some other internetworking
device that keeps track of routes (and, in some cases, metrics associated with
those routes) to particular network destinations.
Serial Line Interface Protocol (SLIP)--Internet protocol used to run IP over
serial lines such as telephone circuits or RS-232 cables interconnecting two
systems. SLIP is now being replaced by PPP.
Simple Network Management Protocol (SNMP)--The Internet network management
protocol. SNMP provides a means to monitor and set network configuration and
runtime parameters.
System Network Architecture (SNA)--IBM's high-level protocol for
communications between its mainframes, peripherals and other equipment.
Transmission Control Protocol/Internet Protocol (TCP/IP)--The common name
for the suite of protocols developed by the U.S. Department of Defense in the
1970s to support the construction of world-wide internetworks. TCP and IP are
the two best-known protocols in the suite. TCP corresponds to Layer 4 (the
transport layer) of the OSI reference model. It provides reliable transmission
of data. IP corresponds to layer 3 (the network layer) of the OSI reference
model and provides connectionless datagram service.
Wide Area Network (WAN)--A network which encompasses interconnectivity
between devices over a wide geographic area. Such networks would require public
rights-of-way and operate over long distances.
| 
