QoS license control definition
QoS license control is a component of Windows 2000, which can manage network resources (bandwidth) in subnet levels. For subnet bandwidth management (SBM), it is based on the Internet Engineering Task Group (IETF) standard.
For network administrators, the effective use of bandwidth and its allocation issues are very concerned, especially for real-time, multimedia programs, more worthy of attention. QoS license control solves this problem by preventing programs that exceed the bandwidth exceeding the processing capabilities of the subnet. It controls the number of bandwidths that can be retained, transmit communication, and users who retain priority bandwidth.
You can create and manage QoS license control policies using QoS license control console. QoS license control uses policy-based management to determine the time and mode of allocating priority bandwidth, and determine the user who retains the priority bandwidth. You can configure policies to meet the needs of users, programs, sites, or enterprises.
All subnet users who enable SBM can use QoS license control to request priority bandwidth. This includes all hosts running Windows 2000, Windows 98 or SBM client software. The client that sends data on the subnet must enable QoS. In this document, the "Subnet Customer" represents a customer who enables SBM enabled QoS programs.
QoS license control benefits
Windows 2000 QoS license Control simplifies bandwidth management in a very low cost. you can:
Configure strategies and subnet configurations on the QoS license control console.
Use the identity of the user and subnet to retain network resources and set priority.
Keep the bandwidth reserve transparent to the user without requiring users to receive training.
Separate network resources between low priority and high priority communication.
Protect the end-to-end transport service by ensuring a lower delay.
LAN, WAN, ATM, Ethernet and token ring network configuration interaction.
Support for multicast transmission of bandwidth reserved messages.
Processing bandwidth reserved messages encrypted by a Windows 2000 IP security mechanism.
Service quality (QoS)
To understand the mode of work of QoS license control, first understand "service quality" (QoS).
Typically, the quality of service is a group of methods or processes, and the service-based organization maintains specific quality levels by performing these methods or processes. In a Windows 2000 environment, QoS is a set of service needs, and the network must meet these needs to ensure proper service level data transmission.
The implementation of QoS can make real-time programs use network bandwidth. Because it ensures a guarantee level with sufficient network resources, it provides a service level similar to a private network for shared networks.
QoS guarantees a level of service that allows the program to transfer data in an acceptable manner and acceptable time frame.
The Target of QoS is to establish a guarantee transmission system for network communication, such as an Internet Protocol (IP) packet. This goal (including services and protocols) is implemented by a set of QoS mechanisms. QoS's Windows 2000 implementation includes:
Subnet Bandwidth Management (SBM) service for controlling the bandwidth on the subnet.
The integrated resource retention protocol (RSVP) allows both transmission and reception of both parties to establish a QoS high-speed channel for retaining.
Communication control services, used to distinguish between communications and plan them.
RSVP overview
QoS license control and its subnet customers use resource retention protocols (RSVP) as a message service for priority bandwidth requests. RSVP is a signal transmission protocol that is retained along the data path determined by the network route selection protocol in advance. this agreement:
Support multicast or unicast transmission.
Pass the reserved request to all network components in the communication path.
The reservation is maintained on a network component that supports the computer and the receiving computer (such as routers, computers, or switches).
Directly transfer between routers and switches that do not support RSVP.
For fully guaranteed reservations, each hop must be allowed to keep and physically assign a request for bandwidth. After the reservation is allowed, the hop submit enough resources.
If rejecting, the program will immediately receive a response that the network does not support the number and type of bandwidth, or cannot support the requested service level. The program is used to determine whether to use the best level of delivery to send data or wait and repeat the request later. RSVP is a "soft status protocol" requesting periodically refreshing reservations. Reserved information or retention state is cached in each hop. If the network routing protocol changes the data path, the RSVP automatically installs the reserved state through the new routing. If you do not receive a refresh information, leave the timeout and will be discarded automatically, and the bandwidth is released.
note
Currently, many routers and switches do not support RSVP. In this case, the reserved message is simply through each hop. Request service level end peer and low delay assurance.
Any current routing protocol can work well with RSVP well with RSVP, and RSVP also supports many network layer protocols (including TCP / IP).
RSVP example
RSVP was originally designed to be launched by the recipient because for a large network, the sender started without a better adjustment in the case of various types of senders. With RSVP, each recipient creates a self-reserved reservation, and all differences in the reservation are resolved by RSVP. If the different receivers requests different resources, the sender and the router will simultaneously consolidate the requested maximum value by getting the request.
In this example, the client sends the RSVP PATH and RESV messages between the multimedia server, the server, and the client to establish an appointment so that the server uses the media interactive component to properly process the appropriate priority tag to return data to the client.
The multimedia server sends a PATH message to the QoS license control host.
The QoS license control host approves the request and sends it to the customer. The PATH message is traversed throughout the network and set the PATH status on each hop. Each Path status contains a Path message and the IP address of the previous hop host.
Customer Create a Resv message indicating that the data is received. The Resv message determines the route using the addressing information in the PATH status by reaching the multimedia server.
When the Resv message reaches each hop, the reservation is approved and distributed and distributed. The hop maintenance reservation (ResV) state is notified that the communication control is about to send data.
During data transfer, multimedia servers and customers periodically send Path and Resv messages to keep the reserved state in the appropriate location.
Communication control overview
Communication control is a QoS mechanism, it can:
Reduce delay and latency (cumulative delay) in network communication transmission.
Work with QoS license control and RSVP to meet the service levels and priorities required in bandwidth requests.
Start when you enable QoS client request QoS.
For subnet customers who do not enable QoS license control (QoS must be enabled).
Control those data streams that do not use the RSVP device.
Service level
The service level of each data stream is part of the communication attribute, and is configured in each QoS license control policy. To determine the required service level settings, you should understand the communication mode belonging to two main groups: elasticity and real-time.
"Elastic communication" is easy to change. When only a small amount of bandwidth is available, the transmission speed of the elastic communication will be slower. When there is enough bandwidth, the transmission speed will speed up. The data sender will automatically adjust to the network rate. Elastic communication is typically generated by interactive or oriented procedures, such as batch data transmission.
The "real-time communication" is characterized by the arrival time interval between any two packets at any two packets, which must be very matched. Real-time communication is limited by its restrictions on network conditions, and the delay will significantly reduce clarity. For example, the environment and content can be changed or shortened during audio conference. If the video communication delay is too long, the image cannot be seen. Supports three service levels in QoS license control. "Best Effect" is suitable for elastic communications. The other two levels apply to real-time programs and awarded priority services for these programs. In these two levels, all unreserved bandwidths, or have been reserved but the currently not used bandwidth is available for other communications.
"Best Effect" is a standard operation process of many IP-based networks. It is an unconnected transmission mode suitable for elastic communication. Packet transmission is performed without ensuring a low delay or sufficient bandwidth.
In the case where no load (load is not very heavy or network is not crowded), "controlled load" is similar to the "best effect" service. In a network member, the stream that receives the controlled load service can only be aware of a short latency, even not a delay, or can feel that there is no blocking.
"Guarantee Service" guarantees the maximum delay restriction. This is very useful if each host on the data path is provided, including routers or switches that match QoS and RSVPs.
Obviously, network users can benefit from the choice guarantee service because it increases the quality of the transmission. However, the impact of ensuring communication on the network is very large, so this is not worth it for programs that generate elastic or best effects.
License control overview
Real-time programs typically send data with RTP or UDP protocol. These protocols allow for transmission speeds than the connection-oriented protocol (e.g., TCP), which are confirmed after transmitting each package, thus generating delays. Since RTP and UDP are non-connectivity protocols, the reliability of their transmission services is limited. If a network is congested and lost a packet, it will not be recovered.
The performance of the real-time program depends on the low latency service, because severe delay and loss will result in distortion or incomplete images and sounds. To deploy real-time programs with acceptable communication rate, you must ensure that network resources have a certain degree of availability. In addition, resource management must allow real-time communication and traditional communications to coexist on the same network. "QoS License Control" is used with "Windows 2000 Service Quality (QoS)", which is a simple and economical mode to achieve this effect, is especially obvious compared to each communication type using different networks.
You can use "QoS License Control" to centrally specify how to use the shared network resources, users and usage time. After installing and configuring this service, the ACS host will control the bandwidth of the subnet it connects. Any host (subnet client) on the subnet (subnet client) can submit a priority bandwidth request to the "QoS License Control" host. The host will determine if the available bandwidth is sufficient according to the following conditions:
Current status of resource availability on subnet
"QoS license control" user policy permissions
You can set any Windows 2000 Server and you can use the network to be set to the "QoS License Control" host. "QoS License Control" runs in IP network layers, for most common program protocol services, including all transport protocols (TCP, UDP, and RTP) in TCP / IP Protocol clusters.
Running Windows 98 or Windows 2000 clients and servers are automatically configured to request bandwidth with the QoS License Control host. "QoS License Control" also supports clients that run any other operating system (running subnet bandwidth management client software) that enable SBM.
The programs that do not identify QoS cannot interact with "QoS License Control", and cannot receive the best service level from the network.
Definition of ATM
Asynchronous Transfer Mode (ATM) refers to many related technologies, including software, hardware, and connection media. ATM is different from other existing LAN and WAN technology, which is specially designed to support high-speed communication. ATM allows the network to use bandwidth resources with maximum efficiency, and maintain "service quality (QoS)" for users and programs with strict service requirements. The basic component of the ATM is a terminal station, a computer connected to the ATM network, an ATM switch, and a device responsible for connecting the terminal station and ensuring data successfully transmitted.
Asynchronous refers to a fixed channel or channel that can be used in accordance with timed mechanisms or clock synchronization. Devices that communicate with asynchronous communications are not necessarily to transmit and receive information in an accurate transmission rate in accordance with their capabilities. Instead, the sender and the recipient can negotiate their communication rate, but must be based on physical hardware limits, and the ability to maintain reliable information flow within the network.
The transmission mode refers to a method of transmitting information between senders and recipients. In ATM, the concept of small units fixed in length is used to construct and pack data to be transmitted. By using this unit with the variable length packet mechanisms used in most existing network technologies, ATM ensures negotiation and management of the connection, so that a single data type or connection exclusive transmission channel will occur.
ATM service quality (QoS)
In the ATM network, service quality (QoS) is used to ensure that all network users can always achieve good results. When providing services for any connection, ATM is executing QoS through three conventional methods:
When the required connection service quality level is required, ATM will guarantee and force all devices that provide this connection to meet the required service level.
When a connection requires its preferred service quality, the ATM will try to obtain resources from the network to meet the required service level while ensuring and maintaining the quality of service that requires the connection of the resource.
If a connection does not specify a service quality, ATM will try to use the available network resources to provide similar services similar to other LAN / WAN transmission modes.
