QoS General Topics
1. What is the terminology used by QOS?
2. Why should I implement QOS?
3. What are the queuing mechanisms?
4. What are QOS Policies and how do they translate
to business applications?
5. What is Differentiated Services (Diff-Serv)?
6. What is Traffic Shaping / Conditioning?
7. What is Bandwidth Brokering?
8. What is MPLS?
1. How does NeuTrans do network discovery?
2. How can "Adaptive Learning" help manage
QoS on my Network?
3. What are the Fault Tolerant Capabilities of NeuTrans?
4. Will NeuTrans scale with the growth of my network?
5. Does NeuTrans work with all vendor hardware?
6. What are Role definitions?
7. What are 'point bandwidth management solutions'
and how is Asce Networks NeuTrans different?
Development (COPS / API / SDK)
1. What is Common Open Policy Service (COPS)?
2. What is Resource Reservation Protocol (RSVP)?
3. What is the Difference between COPS-PR and COPS-RSVP?
4. What is "Self-Provisioning", and how
can this be integrated into NeuTrans?
1. What is the terminology
used by QoS?
Asce Networks has put together a glossary of terms
to assist you as a resource in getting familiar with QoS terminology.Go
to our Glossary for more info.
Why should I implement QOS?
Quality of Service can encompass many aspects of a
network. People, users and/or customers view network quality as
the process of delivering data in a reliable and efficient manner.
As the reliable delivery of data has moved from a luxury to a necessity,
so has the intelligence of the network. As customer demand increases,
the ability to provide this intelligence has become a focus. The
addition of increased bandwidth is no longer an option. The convergence
of Voice, Video, and Data has necessitated that not only should
the data path be reliable but efficient. Some applications require
not only reliable delivery but also require managed latency. With
Quality of Service this has now become a reality and Asce Networks'
NeuTrans is the tool needed to manage these mechanisms.
What are the queuing mechanisms?
There are typically three methods of
FIFO (First in First Out) - FIFO queuing is a strict queuing method
of transmitting packets in the order in which they are received.
CBQ (Class Based Queuing) - A queuing methodology in which traffic
is divided into classes and separated into queues according to the
WFQ (Weighted Fair Queuing) - WFQ allows multiple queues to be defined
for multiple traffic flows. The administrator creates different
queue sizes and delegates what traffic is destined for a particular
In addition to the 3 basic queuing mechanisms, Asce Networks' NeuTrans
utilizes advanced queuing methodologies for Cisco, Juniper and Nortel
Networks etc. routers that combine characteristics of the basic
types. Examples of these mechanisms are; CQ (Custom Queuing), PQ
(Priority Queuing), CCBQ (Custom Class Based Queuing) and CBWFQ
(Class Based Weighted Fair Queuing).
What are QOS Policies and how do they translate to business applications?
QOS policies are a compiled set of
instructions that define how networking devices should prioritize
and forward network traffic. Business applications can suffer tremendously
when combined with other traffic types. In a best-effort networking
environment, QOS can truly be appreciated when Business Applications
compete with non-business downloads, or a company's CRM application
is battling the morning rush of web browsing. For the Service Provider,
the ability to offer premium services based on flexible quality
levels that are adaptive to the changing demands of users can prove
to be the competitive edge needed.
What is Differentiated Services (Diff-Serv)?
Diff-Serv is a methodology used to
specify and control network traffic so that certain types of traffic
get priority. Diff-Serv is more complex than traffic tagging in
that Diff-Serv has the ability to handle packets according to a
Per Hop Behavior (PHB) that is selected on the basis of the contents
of the DS field in the packet header.
What is Traffic Shaping/Conditioning?
Traffic Shaping or Conditioning is
the enforcement of the rules on network packets based upon QOS policies.
In QOS networks, it is necessary to specify the traffic profile
for a "connection" to decide how to allocate various network
resources. Traffic Shaping/Conditioning ensures that traffic entering
at an edge or a core node adheres to a specified profile.
What is Bandwidth Brokering?
Bandwidth Brokering is the concept
of providing service levels with quantitative, absolute bandwidth
assurance. The service may be provided entirely within one domain,
from domain-edge to domain-edge or across a number of domains. Bandwidth
brokers are designed to be configured with policies within an organization
that monitor current allocation of marked traffic and make decisions
on new requests based on current network resources.
What is MPLS?
MPLS (Multi Protocol Label Switching)
is a technology that overlays a label swapping approach to packet
forwarding on top of the well-understood hop-by-hop routing mechanisms
of IP networks. Fundamentally, it separates the "where"
of packet forwarding from the "how". This technique brings
a number of benefits meeting the needs of business and ecommerce
with IP networks.
A new standard for a new world of networking, MPLS is an emerging
Internet Engineering Task Force (IETF) standard based on Cisco Tag
Switching. MPLS is an innovative approach that uses a label-based
forwarding paradigm. Labels indicate both routes and service attributes.
At the ingress edge, incoming packets are processed and labels selected
and applied. The core merely reads labels, applies appropriate services,
and forwards packets based on the label. Processor-intensive analysis,
classification, and filtering happens only once, at the ingress
edge. At the egress edge, labels are stripped, and packets forwarded
to their final destination.
1. How does NeuTrans do network discovery?
NeuTrans discovers network devices
using two methods, based on whether or not the device support the
Common Open Policy System (COPS). For COPS compliant devices, the
device itself initiates communication with the Policy Server. Non-COPS
devices are discovered via the SNMP. The system administrator enters
one or more SNMP community strings into the Policy Console, and
discovery of devices is based on matched community strings.
How can "Adaptive Policy Technology" help manage QoS on
Adaptive Policy Technology simplifies
Policy administration by detecting the QoS capabilities of different
OS versions. NeuTrans interrogates the router to determine its QoS
capabilities. The system then implements policies based on capabilitiesit
received during the interrogation process. This allows users to
upgrade the OS on their current QoS enabled devices, without necessitating
the upgrade of NeuTrans.
What are the Fault Tolerant Capabilities of NeuTrans?
Every component of the NeuTrans has
fault tolerant capabilities. The flexibility of the Policy Server
allows for configuration as a "Hot-Standby" or "Load
Balancing", where the Policy Servers would share managed devices.
The Policy Administrator can be configured as a Primary or Backup
component. In the event of a failure of the Primary Policy Administrator
the Backup Administrator would assume responsibility as the Primary
Policy Server. The system also supports the autonomous operation
of Policy Servers and could preclude the necessity of the Policy
Administrator. In this error condition the current deployed policies
would be enforce, but new policies could not be deployed.
Will NeuTrans scale with the growth of my network?
Absolutely. Asce Networks NeuTrans
has been designed to handle tens of thousands of devices under management,
through its scalable, distributed architecture. When it comes to
the number of users and applications, the limits of the devices
in the network will be reached well before NeuTrans.
Does NeuTrans work with all vendor hardware?
Asce Networks' mission is to allow
network managers who manage IP networks to deliver stable, predictable
performance from those networks using a powerful, policy-based management
approach. Asce Networks is committed to maximizing the capabilities
of existing network infrastructure, and to that end, we support
multi-vendor networks and existing devices through our open Device
API and SDK, ensuring network managers have a free choice of device
vendors in the future. So, networks can be built with specialist
devices such as traffic shapers and firewalls, safe in the knowledge
that the maximum abilities of these devices can be unleashed by
Asce Networks. Currently, NeuTrans supports Cisco, Juniper, Nortel
Networks, Lucent, Riverstone, Unisphere, Avici etc.
What are Role definitions?
Role definitions allow users to assign particular interface
types and speeds to groups (Roles). These Roles can then be used
in the definition of Policy. This provides for greater ease in the
creation of policy by allowing the user to incorporate these roles,
rather than the individual interfaces, into the Policy definition.
What are 'point bandwidth management solutions' and how is Asce
Networks NeuTrans different?
'Point bandwidth management solutions'
are extremely effective in addressing issues of traffic prioritization
at the access link to a web server farm for instance. Without Asce
Networks NeuTrans and the appropriate device driver, they require
prohibitive levels of effort to deliver differentiated levels of
service on a co-ordinated, network-wide basis in any sizable network.
| Development (COPS
/ API / SDK)
1. What is Common Open Policy Service
COPS is a reliable, responsive protocol
for policy-based network management systems, enabling the dynamic
and proactive management of network devices. By integrating COPS,
hardware manufactures and equipment providers can "policy-enable"
their products, offering complete support for industry standard
policy management systems such as Asce Networks' NeuTrans. By integrating
the standards-based COPS protocol, networking companies can extend
their products to access intuitive, policy-enabled networking solutions.
What is Resource Reservation Protocol (RSVP)?
One of the key IETF protocols that
communicates the QoS requirements for a given application to a device
in the path of the transmission. A reservation for the required
bandwidth is allowed or denied depending on the current network
conditions. By itself, Resource ReSerVation Protocol provides a
way to reserve capacity one device at a time. In a centrally managed
QoS system, Resource ReSerVation Protocol can be implemented according
to policies that apply across the network. Resource ReSerVation
Protocol is expected to be utilized predominantly in the campus-level
RSVP is designed to operate within the framework of current and
future routing protocols. An RSVP process consults the local routing
database(s) to obtain routes. Routing protocols determine where
packets get forwarded; RSVP is only concerned with the Quality-of-Service
for those packets that are forwarded in accordance with a routing
What is the Difference between COPS-PR and COPS-RSVP?
As previously stated, COPS is a reliable,
responsive protocol for policy-based network management systems,
enabling the dynamic and proactive management of network devices.
The differences between COPS-PR and COPS-RSVP are as follows: The
COPS-PR architecture is one of provisioning. The Policy Enforcement
Point (PEP) sends to the Policy Decision Point (PDP) a request with
a list of policy requirements (Queuing, Marking¡ etc.). The PDP
provides the policy in a PDP-decision. Afterwards the COPS-conversation
can be potentially limited to Keep-Alive. The PDP may send unsolicited
policy updates later, and the PEP may issue more requests or update
past requests. In the COPS-PR architecture the Global pre-prepared
policy is used to manage all devices.
The COPS-RSVP architecture is one of outsourcing. There is no global
pre-prepared policy. The PEP requests the PDP for policy action
only when an outside source triggers it for action. The PDP decision
is relatively simple, permit/deny. As an example the RSVP flow receiver
(the outsource) sends a reservation request to the PEP (a router),
asking for a certain bandwidth for the RSVP Flow. The PEP forwards
the request to the PDP. The PDP decision might be: accept / reject
/ accept with changes. To summarize, in the provisioning model,
all QOS is pre-prepared and dynamic provisioning is managed and
allocated by the PDP. In the outsourcing model the QOS is not provisioned
until an outside source (outsourcing) requests QOS.
What is "Self-Provisioning", and how can this be integrated
Self-Provisioning is a methodology
that allows customers / clients to select and deploy QoS mechanisms.
Part of Asce Networks' overall solution is the ability to have customers
/clients provision network bandwidth and QoS without the need for
IT / IS intervention.
From a Service Providers perspective they could have available 4
classes of service (Bronze, Silver, Gold, and Platinum). The customer
decides to upgrade from their current class of service (Bronze)
to Gold. The customer would supply customer information and then
select the appropriate service (Gold). The underlying functions
would then provide the appropriate bandwidth / QoS, and adjust billing
records as appropriate. From an enterprise perspective, it can be
used when the CEO wants to do video conferencing. The CEO can select
the appropriate attendees, conference length, and the network will
be provisioned accordingly.
This document provides brief definitions
for key Performance Provisioning and Quality of Service (QoS) terms.
These terms are used on the Asce Networks' Web site and throughout
Adaptive Policy Technology
A component of the Asce Networks NeuTrans Policy Server. Adaptive
Policy technology allows Policy Servers to dynamically adapt to
changes in the network infrastructure. QoS capabilities are highly
dependent upon vendor type, model number, technology type, and software
version. The Adaptive Policy technology automatically learns this
information and ensures that NeuTrans will dynamically adapt to
Any separately managed network, whether departmental, divisional,
or company-wide. There are several different kinds of domains, such
as NT domains and IP domains. For quality of service enforcement
purposes, a network domain refers to any domain that shares a common
QoS policy. Usually managed by a single corporate entity.
Admission Control Any technique for controlling the admission of
network traffic from outside a given administrative domain.
Transfer Mode (ATM)
A data framing and transmission architecture designed to carry voice,
video and data, which has built-in QoS capabilities. Operates at
Layer 2 of the OSI model. Although relatively few native ATM applications
exist, TCP/IP traffic can be sent over an underlying ATM layer.
This practice is now prevalent for the implementation of Wide-Area
Networks and the Internet backbones operated by major telecommunications
companies. ATM-based technology is also finding its way into systems
for improving TCP/IP connectivity on corporate networks. Although
generally considered too costly to be an end-to-end (desktop-to-desktop)
QoS solution, ATM services are likely to co-exist with QoS-enabled
IP networks for many years to come.
The superstructure of the Internet where the national Internet Service
Providers link directly to each other. Sometimes used to refer to
the trunkline of any network. Usually the backbone speed is significantly
higher than that of the networks that it connects. A connecting
network highway for major networks including large enterprise networks
and the Internet.
A measure of data transmission capacity, usually expressed in kilobits
per second (Kbps) or megabits per second (Mbps). Bandwidth indicates
the theoretical maximum capacity of a connection, but as the theoretical
bandwidth is approached, negative factors such as transmission delay
can cause deterioration in quality. If you increase bandwidth, you
can transfer more data. Network bandwidth can be visualized as a
pipe that transfers data. The larger the pipe, the more data can
be sent through it.
A rudimentary traffic management solution deployed at congestion
points that limits access to network resources (e.g., it might prevent
Pointcast-like "push" traffic from entering the network).
Because a bandwidth manager solution often requires locating a proprietary
hardware device directly on the network, it may be an additional
point-of-failure. Not an end-to-end solution, a bandwidth manager
cannot coordinate multiple traffic flows or resolve conflicting
QoS requests made by multiple clients. See also PointQoS.
The default behavior of TCP/IP networks in the absence of QoS measures.
TCP/IP nodes will make their best effort to deliver a transmission
but will drop packets indiscriminately in the event of congestion.
A methodology for classifying packets and queuing them according
to criteria defined by an administrator. The queuing system is designed
to prevent any one application from monopolizing the system. Also
known as Custom Queuing. See also Weighted Fair Queuing.
of Service (CoS)
A category based on type of user, type of application, or some other
criteria that QoS systems can use to provide differentiated classes
An attempt to head off congestion before it can occur. Random Early
Detection (RED) is an example. See also Congestion Management.
A mechanism that imposes order when traffic exceeds network capacity.
It determines whether some packets must be discarded, and, if so,
it preserves the more important packets. Queuing, scheduling, and
traffic shaping are among the most popular techniques. See Also
A high level but not guaranteed service. In the proposed IETF Integrated
Service model, this level of service is designed for multimedia
applications where time delay is not critical but quality of the
delivery is important. This service is appropriate for applications
such as one-way voice or video, but not for real-time applications,
such as teleconferencing. See also Guaranteed Service.
A network that combines varied traffic types such as data, voice,
and multimedia. Most analysts expect the converged network of the
future to be based on Internet protocols. This trend is evident
in corporate networks, which are starting to combine videoconferencing
on their traditional data networks, as well as in the merging of
the telephone, cable television, and Internet service industries.
Open Policy Service (COPS)
An IETF proposed standard for implementing QoS policies as an end-to-end
service. Common Open Policy Service allows a policy server to control
the devices on the network, such as routers and switches, so that
a cohesive policy based on business priorities can be achieved.
Common Open Policy Service is a companion protocol to Resource ReSerVation
Protocol. The QoS policy exchange mechanism. See IETF Internet Draft:
See Class-based Queuing
An emerging Quality of Service standard. A superset of IP Precedence/CBQ.
By utilizes an entire TOS byte in the IP header, it offers up to
256 levels of priority. This protocol is expected to be used predominanty
in the IP backbone environments. See IETF Internet Draft: http://ietf.org/internet-drafts/draft-ietf-diffserv-arch-01.txt
A device such as a router or a gateway that is deployed at the border
of an administrative domain. Such devices control traffic through
one point only. Contrast with End-to-End QoS.
A system that enforces consistent Quality of Service policies throughout
a network. Has the ability to provide both class of service and
reserved bandwidth for different types of network traffic. End-to-end
QoS coordinates and enforces predefined traffic management policies
across multiple network devices.
A service level that attempts to guarantee a minimal delay for traffic
delivery. In the proposed IETF Integrated Service model, guaranteed
service is intended for real-time applications, such as teleconferencing.
The guarantee is not absolute, but such traffic is a level above
controlled-load service. See also Controlled-load Service
802. 1p Standard
An IEEE standard that governs the prioritization of packets in the
Ethernet and Token Ring networks. Offers eight discrete priority
levels., ranging from the default of best effort, through excellent
effort (a business-critical application, but tolerant of some delay),
interactive multimedia (sensitive to delay or jitter), and reserved
(highest priority). This is a Layer 2 (Data Link) priority setting,
as opposed to the ToS and IP Precedence/CBQ bits, which are Layer
3 (Network Level) settings carried in the IPv4 header. Because it
must be implemented in the hardware of network devices, existing
switches and routers need to be replaced with ones supporting this
technology. See also IP Precedence/CBQ, Type of Service.
The practice of supporting audio, video, and real-time data within
a single network infrastructure. The IETF's Integrated Services
(IntServ) working group has proposed a variety of QoS standards
to support such systems. See also Controlled-load Service, Guaranteed
Service. See IETC RFC 1633 http://info.internet.isi.edu/in-notes/rfc/files/rfc1633.txt
and 2210 http://info.internet.isi.edu/in-notes/rfc/files/rfc2210.txt
Engineering Task Force (IETF)
The international organization that defines Internet protocols and
standards, including refinements to improve QoS.
Service Provider (ISP)
A telecommunications company that provides dialup or leased-line
connections to the Internet. Local and regional ISPs forward traffic
to backbone ISPs, the large carrier-class ISPs that own the national
and international Internet backbone infrastructure. Many ISPs offer
additional services, such as Web site hosting, Voice over IP (VoIP),
or virtual private networks (VPN).
A proactive, business-policy driven, IP traffic management solution
that utilizes Policy Based Network Management (PBNM) technology
to optimize network performance. IP Performance Provisioning simplifies
network traffic management and complex configuration issues by automating
labor intensive, error-prone configuration tasks, affecting many
network devices with one policy rule.
IPv4 (Internet Protocol)
The most widely deployed version of the Internet Protocol, IPv4
provides some basic traffic classification mechanisms with its IP
Precedence/CBQ and Type of Service header fields. However, network
hardware and software traditionally have not been configured to
An update to the Internet Protocol that is in the early phases of
adoption. Most of the refinements concentrate on basics such as
expanding the IP address numbering scheme to accommodate the growth
of the Internet. However, IPv6 does include a Class header field
that is explicitly intended to designate a Class of Service (an
extension of IPv4's IP Precedence/CBQ field).
A technique for making a single transmission fan out to multiple
recipients. Instead of sending a copy of each packet to multiple
destinations, the sender transmits one packet to a multicast group
address. Specialized multicast systems then retransmit each packet
to the individuals in the group. IP Multicast makes more efficient
use of bandwidth by minimizing duplication, particularly in multipoint
applications such as broadcast video traffic (e.g., distance learning).
One of the key QoS transmission standards. A 3-bit value in the
IP packet header meant to designate the relative priority of a packet.
Offers eigth levels of priority from 0 to 7. For example, a brokerage
firm might assign a higher IP Precedence/CBQ value to real-time
stock trades than to e-mail to ensure that the trading gets expedited
delivery. Same as TOS bits - see IETF RFC 1349
A type of distortion that is caused by packets arriving at irregular
intervals. This distortion is particularly damaging to multimedia
traffic. For example, the playback of audio or video data may have
a jittery or shaky quality.
The Physical network layer in the Open Systems Interconnection model.
Applies to basic wiring. Examples of Layer 1 protocols include 10BaseT,
V.35, and AAL1.
The Data Link layer in the Open Systems Interconnection model. Examples
include Ethernet, Token Ring, and HDLC.
The Network protocol layer in the Open Systems Interconnection model.
Examples include IP and IPX.
A traffic-shaping mechanism in which only a fixed amount of traffic
is admitted to the network. Excess traffic is held in a queue until
either it can be accommodated or must be discarded. The analogy
is with water flowing into a leaky bucket. If the water continues
to flow in faster than it can leak out the bottom, the bucket eventually
overflows. See also Token Bucket
Access Protocol (LDAP)
A standard for repositories that store user profiles and other information
about the network. LDAP repositories make it easier for users to
find the e-mail addresses and public key encryption codes of other
users. Currently, there is no stanadard governing storing the QoS
policy information in the LDAP format. However, such suport is expected
in the future. LDAP repositories are not capable of issuing dynamic
policy authorizations and do not have policy enforcement capabilities.
An expensive approach to addressing current limitations of "best
effort" networks by provisioning more bandwidth than expected
network peak requirements. over-provisioning increases the probability,
but does not guarantee the quality, of transmission of time-sensitive
and bandwidth-intensive applications.
A reciprocal agreement that lets Internet Service Providers share
backbone links, so traffic can reach destinations beyond the ISP's
PointQoS offers proprietary hardware-based solutions that address
traffic flow congestion on a single point in the network, usually
at the network edge. Based on the negative principle of blocking
and filtering specific data types or applications (e.g., PointCast),
PointQoS introduces an extra point of failure, does not guarantee
the quality of the data transmission across the entire data path,
and does not address the needs of time-critical applications.
A component of the Asce Networks NeuTrans suite. It is a centralized
repository that maintains all QoS policies for the entire network.
Automatically distributes up-to-date policy information to all Asce
Networks NeuTrans Policy Servers and provides failover protection
in the event a NeuTrans Policy Server stops functioning.
A component of the Asce Networks NeuTrans Open Policy System. It
is a server that authorizes QoS requests received from Common Open
Policy Service-enabled routers or Policy Gateways and coordinates
bandwidth usage on multiple network devices to ensure consistent
end-to-end service throughout the data-path. The Policy Server ensures
that packets receive the appropriate Quality of Service, based on
a set of policies defined by the network administrator.
Quality of Service
Quality of Service is the network's ability to match user and application
requirements to network capabilities. It is based on a set of intelligent
network protocols and services used to efficiently control the movement
of information through local or wide area networks. QoS software
sorts and classifies IP packet requests into different traffic classes
and allocates the proper resources to direct traffic based on various
criteria including application type, user or application ID, source
or destination IP address, time of day, and other user-specified
Any system for transmitting QoS requests and parameters between
devices or applications. Resource ReSerVation Protocol is an example
of a QoS signaling system.
A method for metering the flow of traffic by placing packets in
holding queues, and retransmitting them according to a sorting algorithm,
typically a simple first-in-first-out (FIFO) formula. Queues of
different sizes can be used to assign levels of importance according
to Class of Service designations. Queues that overflow typically
discard packets to reduce network congestion.
Early Detection (RED)
A basic congestion avoidance technique built on the base-level TCP
behavior of automatically slowing transmissions when packet loss
is detected. RED tries to anticipate congestion by monitoring a
queue. When the specified threshold is reached, it randomly discards
packets. This is an implicit signal that the originating applications
should slow their transmissions before congestion becomes severe.
See also Weighted Random Early Detection
ReSerVation Protocol (RSVP)
One of the key IETF protocols that communicates the QoS requirements
for a given application to a device in the path of the transmission.
A reservation for the required bandwidth is allowed or denied depending
on the current network conditions. By itself, Resource ReSerVation
Protocol provides a way to reserve capacity one device at a time.
In a centrally managed QoS system, Resource ReSerVation Protocol
can be implemented according to policies that apply across the network.
Resource ReSerVation Protocol is expected to be utilized predominantly
in the campus-level networks. See also Common Open Policy Service.
See IETF RFC 2205 http://info.internet.isi.edu/in-notes/rfc/files/rfc2205.txt
(Service Level Agreement)
An agreement between a business or residential customer and a network
service provider or between two independent service providers that
specifies the minimum service levels that must be provided. Cooperative
service level agreements can be written to ensure consistent handling
of differentiated traffic flows across multiple network domains,
allowing end-to-end quality of service for Internet traffic.
(Transmission Control Protocol/Internet Protocol)
The standard protocol suite for the Internet and an increasing number
of corporate networks. IP is the base network protocol (on which
multiple transport protocols have been implemented) and TCP controls
the behavior of packet transmission.
A technology implemented at network end points that attempts to
regulate the introduction of traffic into the network.
A traffic-shaping mechanism in which a predetermined amount of tokens
in a bucket represent the capacity allowed to each class of traffic.
Packets are forwarded until they exhaust their supply of tokens.
When the token supply is exhausted, packets may be discarded or
delayed until the bucket is replenished. In some systems, a customer's
token supply might correspond to a service fee. See also Leaky Bucket.
A group of techniques that attempt to regulate or meter the flow
of packets through the network. See also Leaky Bucket, Token Bucket
of Service (ToS)
A 4-bit value in an IP packet's message header that identifies the
type of application generating a given traffic flow. Like the IP
Precedence/CBQ field, the ToS value can be used for traffic classification.
Functionally, TOS and IP Precedence/CBQ are identical.
Private Network (VPN)
A way of duplicating the security and reliability of a dedicated
network connection over a less expensive Internet link. This requires
that each network participating in the VPN deploy a compatible firewall
for encrypting messages and for permitting authorized access from
remote locations. A combination of VPN and ThruQoS offers guaranteed
quality and reliability of traffic flow and allows for the migration
of customers from dedicated lines to shared IP infrastructure.
over IP (VoIP)
An emerging technology for carrying phone conversations over the
Internet and intranets. VoIP is sometimes provided as part of a
package with other collaborative applications, such as text-based
chat. ThruQoS is important for VoIP because the audio signal must
come through in a steady stream, just as it would over a traditional
Fair Queuing (WFQ)
A methodology for segmenting traffic into multiple queues, giving
greater weight to certain traffic types by assigning larger queues.
Like class-based queuing (CBQ), WFQ is designed to prevent any one
traffic type from entirely eclipsing another. By default, WFQ favors
lower-volume traffic flows over higher-volume ones (for example,
a routine e-mail over a large FTP download). See also Class-based
Random Early Detection (WRED)
A congestion avoidance technique that takes advantage of TCP's interpretation
of packet loss as a sign to slow transmissions. WRED monitors a
queue until it fills to a specified threshold. It then begins discarding
packets, starting with those that have the lowest IP Precedence/CBQ.
See also Random Early Detection.