July 1999
The Role Of Layer 3 Switching In Telephony Grade
Campus/ LAN Infrastructures
BY TONY RYBCZYNSKI
Internet Protocol (IP) telephony provides a way to consolidate traffic in the WAN,
resulting in significant savings. More importantly, it opens up the door for new types of
clients leveraging Internet, intranet, and LAN connectivity, and enables extensions to
existing applications and a new range of IP-enabled applications. Whether Internet
telephony is addressed as an evolution of the existing voice communications environment or
as an application added on top of an IP network, the IP networking infrastructure has to
provide the reliability and quality of service (QoS) expected by telephony users. In the
campus environment, Layer 3 routing switches supporting high levels of reliability,
multi-gigabit optical networking, and policy-enabled networking are the fundamental
infrastructure building blocks required to support Internet telephony.
The key requirements placed on the campus networking infrastructure to support IP
telephony applications fall into the following categories:
- Scalable platforms with competitive price/performance to accommodate voice and data
traffic demands, both at the workgroup and campus backbone levels.
- Optional switch redundancy at the workgroup and campus backbone levels with networking
options that provide fast recovery from failures.
- Differentiated application networking to allow networks to meet application and user QoS
and security needs across the campus and into the WAN.
- Simplified network management that supports the objectives for a high-quality
infrastructure and operational practices compatible with business critical operation.
Todays installed base of campus network infrastructures is seen as a bottleneck
in meeting the infrastructure needs of Internet telephony, by not being able to
differentiate between voice and data applications (without a significant performance hit),
and not being reliable enough without costly equipment duplication. Campus networks vary
in size from a couple of workgroups to hundreds of workgroups handled by multiple campus
switches. Campus networks today consist of a complex multi-tiered network, e.g.:
- The access or workgroup tier;
- The campus distribution tier;
- The campus core tiers; and
- The server aggregation tier.
This mixture of shared media LAN hubs and switches, multilayer switches, and/or routers
impedes the scalability of the solution. Often the network deployment has been driven by
minimizing the price per user, with little consideration for the reliability implications
of adding business-critical applications, such as telephony. This complexity and the lack
of affordable switch and network redundancy make outages common, resulting in a network
that is orders of magnitude less reliable than required for telephony. While bandwidth is
relatively inexpensive in the in-building environment, the unpredictable nature of
congestion conditions makes the best effort handling of packets unacceptable for
high-quality telephony. The networking design principles that work very well for TCP/IP
data applications are inadequate for Internet telephony. Finally, in many enterprises,
network management systems are optimized for traditional data traffic rather than for
multimedia and reliability-critical operations.
LAYER 3 ROUTING SWITCHES: THE HEART OF THE CAMPUS NETWORK
The core of the vision for next generation campus networks is a platform,
delivering on the infrastructure requirements of Internet telephony and business critical
data applications, without compromising high performance even when providing extensive
classification and security functionality. This platform is a routing switch that can also
be configured as a Layer 2 switch having Layer 3 and up application-awareness. It can be
deployed in a considerably simplified and more resilient two-tier campus network topology
consisting of the access/workgroup tier and a core switch tier. This vision of next
generation campus platforms eliminates the need for a distribution layer between the
access and core layers, as well as the server aggregation layer between the core layer and
server farms. The overall benefits include lower cost, lower latency, and higher
affordable availability. Platform consolidation can be extended to incorporate switch
server functionality, integrated IP telephony call server capabilities, and interworking
with wireless LAN systems.
Scalability
This consolidated platform scales to hundreds of Gbps and millions of packets per
second, with trunking at speeds of tens of Gbps. The transmission infrastructure is based
on twisted pair to the desktop (complemented by wireless technologies), and single and
multimode fiber in the riser on backbone links and to high-capacity servers.
Current fiber distributed data interface (FDDI) and 100 Mbps campus and metropolitan area
network (MAN) links running over dedicated fiber can be upgraded to gigabit Ethernet over
distances as large as 50 km. Gigabit Ethernet running multilink trunking (MLT) effectively
increases trunk capacity to N Gbps (N up to 16). Ten-Gbps Ethernet is also going to be
available, for example, based on OC192c components from SONET (much the same way gigabit
Ethernet borrowed technology from fiber channel). The next plateau is running
IP on SONET and on Dense Wavelength Division Multiplexing (DWDM) over this fiber.
Switch and Network Resilience
This platform cost effectively supports power, interface, control and switching
fabric redundancy, and hot swappability as required of an infrastructure supporting
Internet telephony. Resilience is provided at Layer 1 through SONET and DWDM features in
the extended campus or MAN. At Layer 2, mechanisms such as MLT provide instantaneous
recovery from failures. At Layer 3, resilience is provided through dynamic routing
protocols such as OSPF, complemented by Equal Cost MultiPath (ECMP) routing and Virtual
Router Redundancy Protocol (VRRP). Multigigabit optical networking not only provides the
bandwidth required by enterprises but also, through close integration with switching and
routing technologies, is the basis of new levels of resiliency required by enterprise
networking. A key capability is the support of redundant MLT at all trunking levels across
the campus. Redundant MLT spreads MLT trunks across multiple switch interface cards,
enhancing reliability and very importantly, allowing quick recovery from trunk and
card failures compatible with the needs of telephony applications.
Differentiated application networking
A comprehensive set of IP QoS capabilities are provided across the campus, operating
under a policy management framework. In an ideal world, all applications would indicate
their requirements at Layer 3 and there would be one QoS standard. Unfortunately, there
are a number of competing standards. For example, some applications will use the Type of
Service (TOS) bits in the Layer 3 IP header as specified in the Differentiated Services
(DiffServ) architecture. Others will use Resource reSerVation Protocol (RSVP) Layer 3
signaling as specified in the Integrated Services (IntServ) architecture. Yet others may
indicate their requirements in the MAC header using IEEE 802.1p, these being mapped mostly
to DiffServ as soon as they leave the local LAN. However, most current applications
cant do this. Application awareness is being built into intelligent workgroup and
campus core levels. Technically, this is also referred to as deep packet filtering,
whereby the Layer 2 or 3 switch examines the received packet header fields beyond those
associated with Layer 2 or 3 respectively, to ascertain which preferential treatment is
called for in the network. For example, an application-aware switch can examine TCP port
numbers as discussed above, or fields of a Layer 5 Real-Time Protocol (RTP) header to
detect the start of an Internet telephony call and ensure appropriate treatment across the
network. Adding Layer 3 intelligence and application awareness at the edge of your network
without incurring a price/performance penalty is very important to achieving
the key objective of being able to offer preferential treatment for certain applications
(such as IP telephony), and end users who cant signal their needs to the network at
Layer 3.
Unified management
Unified management consists of three key components: Network management, policy
management, and service level management. Network management includes performance and
fault management capabilities that can significantly enhance network reliability as
perceived by telephony users. Configuration management, including integrity checks, can
avoid configuration errors that result in loss of logical connectivity that would impact
telephony and data users alike. Providing preferential treatment for certain applications
and users is a key requirement for IP telephony, and is provided in next generation campus
infrastructures, through the addition of switch and network level QoS and security
capabilities under policy management.
Policy management provides a structure of network-wide control mechanisms that ensure
that the right applications and end users have access to network resources.
Policy management is an implementation of a set of rules or policies which dictate the
access and use of resources on a per user, application, or company basis to meet
established business objectives. It is essentially focused on providing end-to-end QoS
(bandwidth, latency, priority) and security (authentication, authorization, auditing).
Policy-enabled networking ensures that applications such as voice, e-commerce, supply
chain management and Web access are given the appropriate treatment. It also ensures that
the highest availability (even under failure conditions) is provided to business critical
applications, simplifies operations by providing a unified directory environment, and
generally lowers the total cost of ownership by making the best use of available
bandwidth.
Finally, service level management is a set of client and management capabilities that
allow the IT manager to proactively track the performance of the network from the end user
application perspective.
TOWARDS TELEPHONY-GRADE CAMPUS NETWORKS
IT managers need to develop infrastructures that can support a broad range of
applications. With the increased focus on e-business applications (including e-commerce,
enterprise resource planning, and e-customer care), all of which are business critical,
there is a growing emphasis on the same level of infrastructure reliability that is
required for IP telephony. For example, in a comprehensive supply chain management
application environment, a customer query with a response time requirement of three
seconds may necessitate a large number of back-office network transactions (e.g., to the
factory, to accounting, to inventory databases). Due to the cumulative nature of delays,
individual transactions may have latency requirements below 100 ms (in the same order of
magnitude as voice). Enterprises need solid networking and management infrastructures that
cost-effectively support this breadth of needs and allow switch and network redundancy to
be deployed where it is needed. Scalable price/performance, optional switch and networking
redundancy, application-optimized networking, and simplified management are the key
requirements of business-grade networking infrastructures for business-critical
applications such as IP telephony and e-business.
In the campus environment, IT managers need to invest in routing switches with scalable
performance even when classification and security features are turned on. These switches
should incorporate resiliency at both the switch and trunking level. They need to support
comprehensive IP QoS and traffic management capabilities under network, policy, and
service level management. Next generation campus infrastructures based on routing switches
deliver three key benefits for enterprise users:
- Business-critical availability.
- Operational simplicity and application optimized performance.
- Lower total cost of ownership.
Partnering with vendors who are industry leaders in both telephony communications and
applications, and in routing switches is a key strategy in meeting the challenge of
building a telephony-grade networking infrastructure. c
Tony Rybczynski is director of strategic marketing and technologies for Nortel
Networks Enterprise Solutions. This business unit offers a full range of enterprise
terminal, workgroup, campus, and wide-area unified networks and applications, through
direct and indirect channels. For more information, visit the companys Web site at www.nortelnetworks.com. E-mail questions or
comments to the author at [email protected].
|