How to open the QoS package scheduler. QoS service myth

One of the most popular destinations in today's work with networks is to reduce voice and video in traditional data networks. One of the problems of this kind of information is that for proper work Video and audio data packets must be transmitted to the recipient quickly and securely, without interruptions or too long delays. However, at the same time, this type of traffic should not prevent the transfer of more traditional data packets.

One of possible solutions This problem is QoS. QoS, or service quality, is technology assigning priorities to data packets. QoS allows you to transmit temporary sensitive packages with a higher priority than other packages.

QoS is the industry standard, and not a standard belonging to Microsoft. However, for the first time Microsoft introduced this QoS standard in Windows 2000. The QoS version from Microsoft has evolved quite strongly since that time, but still meets the industry standards.

In Windows XP Professional, QoS first works as a bandwidth reservation mechanism. When the QoS is enabled, the application is allowed to reserve up to 20% of the total network bandwidth provided by each network adapter of the machine. However, the number of network bandwidth reserved can be adjusted. I will show you how to change the amount of bandwidth redundant in the third part.

To see how the backup bandwidth is used, suppose you have a video conferencing application that requires a priority bandwidth for proper operation. Supposed that this application includes QoS, we can say that it reserves 20% of the entire bandwidth of the machine, leaving 80% of the bandwidth for the rest of the network traffic.

All applications, except for video conferencing applications, use the best delivery technology (Best Effort Delivery). This means that the packets are sent with the same priorities "The first delivered package is served first." On the other hand, the video conferencing applications will always have a higher priority compared to the rest of the traffic, but the application will never be allowed to consume more than 20% of all bandwidth.

However, only the fact that Windows XP leaves part of the bandwidth for priority traffic, does not mean that applications with conventional priority will not be able to use backup bandwidth. Although video conferencing applications enjoy the advantages of higher priorities, reserved bandwidth, the chances of permanent use of such applications are very small. In this case, Windows allows other applications to use a backup and non-backup bandwidth for the highest possible delivery until applications for which part of the network bandwidth will not be used.

As soon as the video conferencing application starts, Windows begins to use reservations in forced. But even in this case, the reservation is not absolute. Suppose Windows reserved 20% network bandwidth for video conference application, but this application does not need all 20%. In these cases, Windows allows other applications to use residual bandwidth, but will constantly monitor the needs of an application with a higher priority. If the application requires more bandwidth, the bandwidth will stand out for it maximum value in 20%.

As I said, QoS is the standard industry, and not Microsoft technology. Being so, QoS is used in Windows, but Windows cannot do this work independently. In order for QoS to work, each component of the equipment between the sender and the recipient must support QoS. This means that network adapters, switches, routers and all other devices used should know about QoS, as well as the recipient and sender operating systems.

If you are interested, then you do not need to install some kind of insane exotic network infrastructureTo use QoS. Asynchronous Transfer Mode - ATM - is an excellent network technology to use QoS, since it is a connection-oriented technology technology, however, you can use QoS with other technologies such as Frame Relay, Ethernet and even Wi-Fi (802.11 x).

The reason for which ATM is such an ideal choice for QoS is that it is able to implement bandwidth redundancy and distribute resources at the equipment level. This type of distribution is beyond the framework of Ethernet capabilities and similar network technologies. This does not mean that QoS cannot be used. This only means that QoS should be applied not as in an ATM environment.

In ATM environment, resources are distributed immediately at the level of physical devices. Since Ethernet and other similar technologies cannot allocate resources in this way, the technologies of this type are based on the assignment of priorities, and not on the true allocation of resources. This means that bandwidth redundancy occurs at a higher level of the OSI model. As soon as the bandwidth was reserved, packets with higher priorities are transmitted primarily.

One point that should be taken into account if you are going to apply QoS via Ethernet, Wi-Fi or other similar technologies, is that such technologies do not have compounds. This means that the sender does not have the ability to check the recipient's condition or the state of the network between the sender and the recipient. And this in turn means that the sender can guarantee the sending of packets with higher priorities primarily, but cannot guarantee the delivery of these packages within a certain time. On the other hand, QoS is able to give such a guarantee on the ATM network, since ATM is a connection-oriented technology.

Windows 2000 VS. Windows Server 2003.

Earlier, I said that Microsoft first introduced QoS in Windows 2000, and that this use of QoS has evolved significantly from that time. Therefore, I want to talk a little about the differences between QoS in Windows 2000 and in Windows XP and Windows Server 2003 (in which this standard is used approximately equally).

In Windows 2000, the QoS application was based on the INTSERV architecture, which is not supported in Windows XP or Windows Server 2003. The reason for which Microsoft decided not to use such an architecture, wounded that the underlying API was difficult to use, and the architecture had a problem With scale.

Some organizations are still using Windows 2000, so I decided to give you some information about how the Windows 2000 QoS architecture works. Windows 2000 uses a protocol called RSVP to reduce bandwidth resources. When bandwidth is requested, Windows must determine when packages can be transmitted. To do this, Windows 2000 uses a signal protocol called SBM (subnet bandwidth - Sunbelt Bandwidth Manager) to inform the sender that it is ready to accept the Covenants. The Admission Control Service (ACS) service checks that the effective bandwidth is available and then or provides or rejects the bandwidth.

Traffic Management API

One of the main problems with network traffic priorities is that you cannot assign priorities on the computer based on a computer that generates it. For single computers is the usual way to use multiple applications, and for each application (and operating system) create a separate flow of traffic. When it happens, each flow of traffic should be assigned to the priorities individually. In the end, one application may require a backup bandwidth, while for another application the best delivery is perfect.

Here the TRAFFIC CONTROL API is entered into the game (Traffic Management Program Interface). Traffic Control API is an application programming interface that allows you to apply QoS parameters to individual packages. Traffic Control API works based on the definition of individual traffic flows, and applications different ways QoS control to these threads.

The first thing that makes Traffic Control API, it creates what is known as Filterspec. Filterspec is, in fact, the filter defining what it means for a package to belong to a specific stream. Some attributes used by Filterspec include the IP address of the source and destination of the package and the port number.

Once Filterspec has been defined, the API allows you to create FLOWSPEC. Flaspec defines QoS parameters that will be applied to the package sequence. Some of the parameters defined by FLOWSPEC include the transfer rate (permissible transmission speed) and the type of service.

The third concept defined by the Traffic Control API interface is a stream concept. The stream is a simple sequence of packets that are subject to one FLOWSPEC. Simply put, Filterspec determines which packages will be included in FLOWSPEC. FLOWSPEC determines whether packages will be processed with higher priorities, and the stream is actually the transfer of packets that are subjected to FLowspec processing. All packets in the stream are processed equally.

It should be mentioned that one of the advantages of the Traffic Control API over the Generic QoS API used in Windows 2000 is the ability to use aggregation (association). If the node has several applications that transmit multiple data streams into a common destination, these packages can be combined into a common stream. It acts even if applications use different port numbers, but provided that the IP address of the source and destination is the same.

General Packet Classifier (Generic Packet Classifier)

In the previous section, I talked about the relationship between Flowspec, Filterspec and the flow. However, it is important to remember that the TRAFFIC CONTROL API interface is just an application program interface. Being such, his work is to define and assign priorities to traffic flows, and not the creation of these streams.

GENERIC PACKET CLASSIFIER is responsible for creating threads. As you remember from the past partition, one of the attributes that was defined in FLowspec was the type of service. The type of service, in fact, determines the priority of the flow. Generic Packet Classifier is responsible for determining the type of service that has been assigned to FLOWSPEC, after which it places the associated packages in the queue corresponding to the service type. Each thread is placed in a separate queue.

Qos Packet Scheduler (Package Planner)

The third QoS component you need to know is a QoS package scheduler. Simply put, the main task of the QoS package planner is the formation of traffic. For this package scheduler receives packages from various queues, and then marks these priority packages and flow speed.

As I spoke in the first part of this series of articles, for the correct operation of QoS various components located between the source of the packages and the destination place must support QoS (that is, to know about it). Although these devices should know how to work with QoS, they should also be aware of how to process ordinary traffic without priorities. To make it possible, QoS uses technology called marking.

In fact, there are two types of marking here. QoS Package Scheduler uses diffserv marking, which is recognized by third-level devices, and marking 802.1p, which is recognized by the second-level devices.

Setting up QoS package schedule

Before I show you how labeling works, it should be noted that you will need to configure the QoS package planner so that everything worked. In Windows Server 2003, the QoS package scheduler refers to optional network components, as well as a client for Microsoft or TCP / IP protocol networks. To enable QoS package scheduler, open the properties page of your network connection Servers and check the box next to the QoS Package Planner, as shown in Figure A. If the QoS package scheduler is missing in the list, click the Set button and follow the instructions.

Figure A: QoS package schedule must be enabled before you can use QoS

Another point that you need to know with regard to the QoS Package Planner is that your network adapter must support 802.1p marking for its correct operation. To check your adapter, click the "Configure" button, Figure A, and Windows will display the properties of your network adapter. If you look at the "Advanced" tab on the properties page, you will see different properties that your network adapter supports.

If you look at Figure B, you will see that one of the properties in the list is 802.1q / 1p VLAN Tagging. You also see that this property is disabled by default. To enable 802.1p marking, simply turn on this property and click OK.

Figure B: You must enable 802.1q / 1p VLAN Tagging

You may have noticed in Figure B, that the property you included is associated with VLAN by tagging, and not with batch marking. The reason for this lies in the fact that priority markers are included in VLAN tags. 802.1Q Standard defines VLANS and VLAN tags. This standard will actually reserve three bits in the VLAN package, which are used to record priority code. Unfortunately, 802.1Q Standard never defines what these priority codes should be.

802.1p The standard was created as a supplement to 802.1Q. 802.1p determines the priority marking, which can be enclosed in the VLAN tag.

802.1p Signal

As I spoke in the previous part, the 802.1p signal transmission is carried out at the second level of the OSI model. This level is used by physical devices as switches. Second-level devices that support 802.1p may view the labeling of priorities that are assigned to packets, and then group these packages in separate traffic classes.

In Ethernet networks, the priority marking is included in the VLAN Tags. VLANS and VLAN Tags are defined by 802.1Q standard, which defines the field of three-bit priorities, but does not actually determine how this field of priorities should be used. It is here that 802.1p is entering the game.

802.1p defines various prioritization classes that can be used in conjunction with 802.1Q standard. Ultimately, 802.1Q leaves the right to select the labeling of priorities for the administrator, so technically you do not need to follow the instructions 802.1p, but 802.1p, it seems, is that everyone is chosen.

Although the idea of \u200b\u200busing 802.1p standards to ensure the labeling of the second level, probably sounds like a clean theory, in fact it can be determined using parameters group Policy. Standard 802.1p provides eight different priority classes (vary from 0 to 7). Packages with higher-grade priorities are processed by QoS with a higher delivery priority.

By default, Microsoft assigns the following priority labeling:

But as I mentioned earlier, you can change these priorities, modifying various parameters of group policy. To do this, open the Group Policy Editor and switch to the Console Console Console Configuration of Computer \\ Administration Templates \\ Network \\ Scheduler QoS Packages \\ The value of the second level priorities. As can be seen from Figure A, there are parameters of group policies that correspond to each marking of priorities that I listed above. You can assign your priority marking levels to any of these types of services. However, we should not forget that these group policy settings are valid only for hosts that use Windows XP, 2003 or Vista.

Figure A: You can use the Group Policy Editor to configure the labeling of the second level priorities.

Separate Services (Differentiated Services)

As I explained in the previous article, QoS performs the labeling of priorities in the second and third levels of the OSI model. This ensures accounting of priorities throughout the package delivery process. For example, switches operate at the second level of the OSI model, but routers are usually operating at the third level. Thus, if the packages were used only 802.1p labeling of priorities, then the priorities would be prescribed the switch, but these priorities would be ignored by network routers. To impede this, QoS uses the Differentiated Services Protocol (DiffServ) protocol to assign traffic priorities at the OSI model. Marking DiffServ is included in the IP packet headers using TCP / IP.

The architecture used by DiffServ was originally defined by RFC 2475. However, many architecture specifications were rewritten in RFC 2474. RFC 2474 defines the DiffServ architecture for IPv4 and IPv6.

An interesting point of IPv4 application in RFC 2474 is that even, despite the fact that DiffServ was absolutely overridden, it is still compatible with the original RFC 2475 specification. This means that older routers that do not support new specifications can recognize the assigned priorities.

Current DiffServ Application uses OCTETS Types of Type of Service (TOS) package services for storage of DiffServ values \u200b\u200b(which is called DSCP value). As part of this octet, the first six bits store DSCP value, and the last two bits are not used. The reason why these marks are inversely compatible with RFC 2475 specification is that RFC 2475 required the first three bits in the same OCTET to use in IP seal information. Although DSCP lengths in length are six bits, the first three bits still reflect the IP sequence.

As in the case of marking 802.1p, which I demonstrated earlier, you can configure DiffServ priorities using various group policy settings. Before I show you how, I will present the standard DiffServ priorities used in Windows:

You may have noticed that DiffServ priority marking uses an absolutely different range than 802.1p. Instead of supporting the range of 0 - 7, DiffServ supports the range of labeling of priorities in the range from 0 to 63, while large numbers have higher priorities.

As I said, Windows allows you to define DiffServ labeling priorities using Group Policy Settings. However, it should be remembered that some more advanced routers will assign their own DiffServ values \u200b\u200bto packages, regardless of those values \u200b\u200bthat are assigned to Windows.

Considering this, you can configure the labeling of DiffServ priorities by opening the Group Policy Editor, and clicking on the Console Console Console Configuration \\ Administration Templates \\ Network \\ QoS Package Scheduler.

If you look at Figure B, you will notice that there are two tabs associated with DSCP, which are located under the QoS Package Scheduler. One of these tabs allows you to assign DSCP priorities marking for packets corresponding to FLOWSPEC, and the second allows you to set DSCP priorities marking for inappropriate packages. The actual parameters themselves are similar to both tabs, as shown in Figure C.

Figure B: Windows manages DSCP priority labels separately for packages that match FLowspec, and which do not match.

Figure C: You can manually assign DSCP priorities marking for different types of services.

Diverse Group Policy Settings

If you look at Figure B, you will notice that there are three group policy parameters that I did not say. I wanted briefly mention what kind of parameters and what they do, for those who could be interesting.

The Limit Outstanding Packets parameter is essentially the value of the service threshold. If the number of superior packets reaches a specific value, the QoS will disable any additional bandwidth allocations for the network adapter, until the value drops below the maximum allowable threshold.

The Limit Reservable Bandwidth parameter manages the percentage of overall bandwidth, which applications with QoS support can be reserved. By default, QoS support applications can reserve up to 80% of the network bandwidth. Of course, any part of the bandwidth, reserved, and at the moment not used by QoS applications can be used by other applications.

The Set Timer Resolution parameter controls the minimum time units (in microseconds) that the QoS package scheduler will use for package planning. In fact, this parameter controls the maximum frequency with which the packages may be queued for delivery.

QoS and modems

In this age, almost universal availability of broadband technologies seems strange to talk about modem. However, there are still many small businesses and home users who use modems as a mechanism for connecting to the Internet. Recently, I even saw a large corporation using modem to communicate with satellite offices, which are located in remote places where there is no access to broadband technology.

Of course, the largest problem of using modems is a limited throughput, which they possess. A less obvious, but no less important problem is that users tend to do not change their behavior online when using modem connections. Of course, users may not experience a lot of desire to download large files when connecting to the Internet via the modem, but the rest of the user behavior remains the same as if they were connected through a broadband connection.

Usually users are not too worried about keeping Microsoft Outlook. constantly open, or view the pages until the files are loaded into background mode. Some users also hold the instant messaging system constantly open. The problem with such a type of behavior lies in the fact that each of these applications or tasks consumes a certain amount of Internet connection bandwidth.

To see how QoS can help, let's consider what is happening under normal conditions when QoS is not used. Usually the first application that is trying to access the Internet has the highest rights to use the connection. This does not mean that other applications cannot use the connection, but rather, Windows believes that other applications will not use the connection.

Once the connection is created, Windows begins to dynamically configure the TCP receive window size. TCP Preparation Window Size is the amount of data that can be sent before you expect confirmation that the data was obtained. The larger the size of the TCP receive window, the greater the packages that the sender can transmit before waiting for confirmation of successful delivery.

TCP Preparation Window Size Need to Care Neat. If the TCP preparation window is too small, will suffer efficiency, since TCP requires very frequent adoption confirmations. However, if the TCP receive window is too large, the machine can pass too much data before learning that the problem arose during the transfer. As a result, a re-transmission of a large amount of data is required, which also affects efficiency.

When the application begins to use the Dial-UP Internet connection, Windows dynamically adjusts the size of the TCP receiving window as the packages are sent. The purpose of Windows here is to achieve a stable state, in which the size of the TCP receiving window is optimally configured.

Now suppose that the user opens the second application, which also requires an Internet connection. After it does it, Windows initiates a TCP slow start algorithm, which is an algorithm that is responsible for setting the size of the TCP preparation window to the optimal value. The problem is that TCP is already used by the application that was launched earlier. It affects the second application in two ways. First, the second application requires much longer time to achieve the optimal size of the TCP receiving window. Secondly, the data transfer rate for the second application will always be slower than the transfer rate for the application running forward.

Good news is that you can avoid these problems in Windows XP and Windows Server 2003 by simply launching the QoS package scheduler. After that, the QoS package scheduler will automatically use technology called Deficit Round Robin whenever Windows detects slow connection speed.

The principle of operation of Deficit Round Robin is dynamic creation Separate queues for each application that needs access to the Internet. Windows serves these queues in the form of a cyclic algorithm, which significantly improves the efficiency of all applications that need access to the Internet. If you are interested, Deficit Round Robin is also available in Windows 2000 Server, but does not automatically turn on.

Sharing Internet Connection

In Windows XP and Windows Server 2003, QoS also contributes to the sharing of Internet connection. As you probably know, the joint use of the connection is a simplified option for creating a router based on NAT. The computer to which the Internet connection is connected physically playing the role of the router and the DHCP server for other computers on the network, thereby providing them with access to the Internet through this host. Sharing Internet connection is usually used only in small, peerging networks, in which there is no domain infrastructure. Large size networks typically use routers based on physical devices, or routing and service remote access.

In the above section, I already explained how Windows dynamically sets the size of the TCP receive window. However, such a dynamic configuration can cause problems when using the Internet connection. The reason for this lies in the fact that the connection between computers on the local network is usually relatively fast. Usually such a connection consists of 100 MB Ethernet, or from 802.11g wireless connections. Although these types of compound are far from the fastest, but they are much faster than most Internet connections available in the United States. It is here that the problem lies.

The client computer needs to be interacting over the Internet, but it cannot do this directly. Instead, he uses the host sharing Connections to the Internet as an access module. When Windows calculates the optimal TCP receive window size, it does it, based on the connection speed between the local machine and the Internet Connection Sharing machine. The difference between the amount of data that the local machine can really get from the Internet, and the volume that she thinks can get, based on the connecting speed to the Internet Connection Sharing host, can cause problems. To speak more precisely, the difference in the speed of connection can potentially cause situations in which the data creates backups In the queue connected to the low-speed connection.

The game comes with QoS. If you set the QoS package scheduler on the Internet Connection Sharing node, the Internet Connection Sharing host cancels the TCP receive window size. This means that the Internet Connection Sharing will set the size of the TCP receiving window for local hosts to the same value that they would have in the case of direct Internet connection. This eliminates the problems caused by the mismatch of the network connection rates.

Conclusion

In this series of articles, I told about QoS and how it can be used to form traffic flow into different types Network connections. As you can see, QoS can force the network to work much more efficiently, forming traffic in such a way that it can use the moments of the smallest load on the network, and guarantee the quick traffic delivery with a higher priority.

Brien Posey (Brien Posey)

In the article, consider how to configure reserved bandwidth in Windows 10. By the default of Windows Reserves 20% of the total bandwidth of the Internet.

Yes, yes, Windows 10 operating system reserves a certain percentage of the bandwidth of your Internet connection for the quality of service (QoS).

According to Microsoft:

QoS may include critical system operations, such as updating windows systems, licensing status management, etc. The concept of the reserved bandwidth is applicable to all system programs. Typically, the package planner limits the system to 80% of the connectivity bandwidth. This means that Windows reserves 20% of the bandwidth of your Internet exclusively for QoS.

In case you want to get this reserved percentage of bandwidth, this article is for you. Below will look at two ways to adjust the reserved bandwidth in operating system Windows 10.

NOTE: If you disconnect the entire bandwidth for your system, that is, install it by 0%, it will affect the actions of the operating system, especially on automatic updates.

Denial of responsibility: Further steps will include the registry editor. Errors When editing the registry may adversely affect your system. Therefore, be careful when editing the registry entries and first create a system recovery point.

Step 1: Open Registry Editor (If you are not familiar with the registry editor, click).

Step 2: In the left panel of the registry editor window, go to the next section:

HKEY_LOCAL_MACHINE \\ SOFTWARE \\ POLICIES \\ Microsoft \\ Windows \\ psched

Note:If the section and parameter " Nonbesteffortlimit.»Do not exist just create them.

Step 3: Now on the right pane of the registry section "PSCHED" Find the DWORD parameter (32 bits) named Nonbesteffortlimit.. Double click on it to change its values:

By default, the parameter matters 50 in hexadecimal or 80 in a decimal calculus system.

Step 4: Select the decimal system and set the value equal to the percentage of the required bandwidth.

For exampleif you set the value equal 0 The reserved bandwidth for your Windows operating system will be completely disabled, that is, 0% is equal to 0%. Press the button Ok And close the registry editor.

Step 5: Restart the PC to change the changes.

If you want to configure or limit the backup bandwidth on multiple computers in your organization / in the workplace, you can deploy the appropriate parameter of the Group Policy object.

Step 1: Open the Local Group Policy Editor

Step 2: Go to section: Computer Configuration → "Administrative Templates" → "Network" → "QoS Packs Planner"

Step 3: In the right window, double-click the policy.

By default, this policy is not specified and the system reserves 20% Internet connection bandwidth. You need to enable it, set the parameter "Restrict the redundant bandwidth" value "Inclusive".

In the first part of this series of articles, I told about what QoS makes, and why it is used. In this part, I will continue the conversation with an explanation of how QoS works. As you read this article, please consider that the information presented here is based on Windows Server 2003 using QoS, which differs from the application of QoS in Windows 2000 Server.

Traffic Management API

Here the TRAFFIC CONTROL API is entered into the game (Traffic Management Program Interface). Traffic Control API is an application programming interface that allows you to apply QoS parameters to individual packages. Traffic Control API works based on the definition of individual traffic flows, and the use of various QoS methods for controlling these threads.

General Packet Classifier (Generic Packet Classifier)

Qos Packet Scheduler (Package Planner)

Setting up QoS package schedule

Before I show you how labeling works, it should be noted that you will need to configure the QoS package planner so that everything worked. In Windows Server 2003, the QoS package scheduler refers to optional network components, as well as a client for Microsoft or TCP / IP protocol networks. To enable the QoS package scheduler, open the properties page of your server network connection and check the box next to the QoS package scheduler, as shown in Figure A. If the QoS package scheduler is missing in the list, click Install button and follow the instructions.

Figure A: QoS package schedule must be enabled before you can use QoS

Figure B: You must enable 802.1q / 1p VLAN Tagging

802.1p The standard was created as a supplement to 802.1Q. 802.1p determines the priority marking, which can be enclosed in the VLAN tag. I will tell you about the principle of operation of these two standards in the third part.

Conclusion

In this article we discussed some basic concepts in the Windows Server 2003's QoS architecture. In the third part, I will tell you about how the QoS package scheduler labels packages. I will also tell you how QoS works in a network with low bandwidth.

No one loves when, when loading a web page, they open for a very long time, and the download of files does not occur at that level, on which I would like. Although, when ordering services from the provider, there were clear 20 or even 100 MB / s, and in fact we do not get such a speed.

Of course, this is an explanation. First, the system takes about 20% for its needs, secondly, the browser comes with an answer from DNS servers, the truth is needed for it.

So that there is no, we will now understand how to increase the speed of the Internet several times.

Turn off QoS speed limit

Usually the system has a limit of 20% by speed, although everyone can be different. To increase the speed of the Internet, you must disable this parameter. To do this, we will use local group policies. Unfortunately, this feature is available only on Pro Editors.

Open the "Run" window using a combination Win + R. And in the window that appears, write such a command: gpedit.msc. .

On the left side of the window that opened windows go to the section: Computer configuration – Administrative templates - Net - QoS package schedule – Limit redundant bandwidth.

We find a "limit backup bandwidth" item there. Press it twice and set the parameter to the position. "Included", and then enter the number “0” In the "bandwidth limitation". Click "Apply".

To make sure whether the QoS package scheduler is working with the Network and Common Access Center. You can get there if you click on the taskbar on the Wi-Fi icon, or with a wired connection with the right mouse button. On the left, go to the "Changing Adapter Settings" section. Press the right mouse button on your connection and select "Properties". There must be a parameter «Qos Packet Scheduler»marked with a check mark.

Disable QoS through the registry

With another version of Windows, except Pro, this instruction can come up. Go to the registry, for this we use the combination of Win + R and enter the command regedit..

We go to the next section:

HKEY_LOCAL_MACHINE \\ SOFTWARE \\ POLICIES \\ Microsoft

Here we find the section Windows, click on it right mouse button and create a new section named PSChed..

Go to the created section and right Create a DWORD 32 bit parameter named Nonbesteffortlimit.. We assign the value to this parameter. «0» .

After the work done restart the computer.

Turn off the internet speed limit in

It happens that when using programs requiring Internet, such as torrent clients, there are speed limit functions that can be active with you.

Take for example a torrent client. If you right-click on an active download, then there is an item "Restriction of reception". We guide the mouse and look at it. Active should be mode "Unlimited".

With other torrent clients in the same way. In the other types of programs, you will have to dig and find something similar.

How to increase DNS cache to increase speed?

As many of you know, the DNS cache allows you to store the IP addresses of the resources on which you have already visited, and the re-visiting will use the DNS cache, which will allow you to open the pages much faster. Unfortunately, its volume is not infinite, but it can be increased.

Go! Press Win + R and enter the command to enter the registry - regedit. A window opens where we left should go to this section:

HKEY_LOCAL_MACHINE \\ SYSTEM \\ CURRENTCONTROLSET \\ SERVICES \\ DNSCACHE \\ PARAMETERS

To the right you need to click on the right mouse button on an empty place and create 4 "DWORD" parameters and give them such names - Cachehashtablebucketsize, Cachehashtablesize, MaxcacheTryttlimit., MaxSoacacheTRYTTLlimit..

Each of them should have these values \u200b\u200b(in order to each) - 1, 384, 64000 and 301.

For successful completion, restart the computer.

TCP autoinage - turn off

The system has such a function that can cause the web pages slowly load, and everything is because of its work efficiency with some servers is not very good. Therefore, it will simply turn it off.

To fulfill this task, we need to open command line With elevated privileges and execute such a command there:

Turbo browser mode to accelerate site boot

In many browsers, there is a "turbo mode" function, which speeds up opening pages. So far, it is available on the following popular browsers: Opera and Yandex Browser. For others, you can download special extensions.

In Opera, this feature is turned on if you click in the upper left corner along the Opera button. Find a function Opera Turbo And activate it.

In Yandex browser, this feature is included in the settings - show additional settings. Next to the section "Turbo" set "Always turned on".

Namebench utility to enhance page download

Many providers, especially commercial, always want to save on the equipment. And when you start visiting websites, access to DNS servers (equipment providers) occurs. If it is cheap, then the download speed of the pages you will be very slow. To correct this problem, we need quick DNS servers, and the NameBench program will help us."START BENCHMARK". The program will start testing large number DNS servers and choose the fastest.

When Namebench finds the desired server, it will show its IP address to be prescribed in the settings of your connection.

Updating the firmware of the router

This is the last item, but no less important. If you use a router, the firmware of which is very outdated, then do not wait for him a miracle. Look on the Internet the firmware on your router and find the instructions for its installation, as well as preserve the old, in order to avoid problems.

Here's actually all the methods that can be used on modern windows versions. Although, maybe there is something else, and if it appears, we will not wage it by the party.

There is not a single person who would have read any FAQ on Windows XP at least once. And if so, everyone knows that there is such a harmful Quality of Service - abbreviated QoS. When you configure the system, it is highly recommended to disable it, because it defaults the network bandwidth by 20%, and as if this problem exists in Windows 2000.

These lines are:

Q: How to completely disable the QoS service (Quality of Service)? How to configure it? Is it true that it limits the speed of the network?
A: Indeed, by default, Quality of Service reserves 20% of the channel bandwidth (any - at least a modem at 14400, at least a gigabit Ethernet). Moreover, even if you delete the QoS Packet Scheduler service from Properties connections, this channel is not released. You can free on the channel or simply configure the QoS here. Run the GROUP Policy applet (gpedit.msc). In Group Policy, we find Local Computer Policy and click on Administrative Templates. Select Network - Qos Packet Sheeduler. Turn on Limit Reservable Bandwidth. Now we reduce the Bandwidth Limit 20% to 0% or simply turn it off. If you wish, you can also configure other QoS parameters. To activate the changes made, it remains only to reboot.

20% is, of course, a lot. Truly Microsoft - Mazdai. Approval of this kind rummage from the FAQ in the FAQ, from the forum to the forum, from the media in the media, are used in all sorts of "twikals" - "Configuring" programs for Windows XP (by the way, open "Group Policies" and "Local Safety Policies", And no "twikalka" will not compare with them on the wealth of setting options). To expose the unscrupulous statements of this kind, it is necessary to be careful that we now and apply a systematic approach. That is, thoroughly study the problematic question, based on official sources.

What is a network with high-quality service?

Let's take the following simplified definition of the network system. Applications are launched and operated on hosts and exchange data from each other. Applications send data to the operating system for transmission over the network. Once the data is transferred to the operating system, they become a network traffic.

The QoS network service relies on the network ability to process this traffic so that you are guaranteed to query some applications. This requires the existence of a fundamental mechanism for processing network traffic capable of identifying traffic that has the right to special processing and the right to manage these mechanisms.

QoS functionality is designed to satisfy two network subjects: network applications and network administrators. They often have disagreements. The network administrator limits the resources used by the specific application, at the same time the application tries to capture as many network resources as possible. Their interests can be agreed, taking into account the fact that the network administrator plays a primary role in relation to all applications and users.

The main parameters of QoS.

Various applications have different requirements for processing their network traffic. Applications are more or less tolerant to delays and traffic losses. These requirements have been applied to the following parametersassociated with QoS:

Bandwidth (bandwidth) - the speed with which the traffic generated by the application must be transmitted over the network;
Latency - a delay that the application can be allowed to deliver a data packet;
Jitter - change the delay time;
Loss (Loss) is the percentage of lost data.

If endless network resources were available, then all the traffic traffic could be transferred at the required speed, with zero delay time, zero change in the delay time and zero loss. However, network resources are not limitless.

The QoS mechanism monitors the distribution of network resources for the application traffic to perform the requirements for its transfer.

Fundamental QoS resources and traffic processing mechanisms

Networks that bind hosts use a variety network devices Including network adapters of hosts, routers, switches and hubs. Each of them has network interfaces. Each network interface can accept and transmit traffic at a final speed. If the speed with which traffic is directed to the interface is higher than the speed with which the interface transmits traffic further, then overload occurs.

Network devices can handle the overload status by organizing traffic queue in the device's memory (in buffer) until the overload passes. In other cases, network equipment may refuse traffic to facilitate overload. As a result, applications are faced with changing the waiting time (since traffic is stored in queues on interfaces) or with traffic loss.

The ability of network interfaces to forward traffic and the availability of memory to save traffic in network devices (as long as the traffic can be sent further) constitute fundamental resources required to provide QoS for application traffic flows.

Distribution of QoS Resources on Network Devices

QoS supporting devices reasonably use network resources for traffic transmission. That is, the traffic of applications, more tolerant to delay, becomes queue (stored in the buffer in memory), and the traffic of applications, critical to delays, is transmitted further.

To perform this task, the network device must identify traffic by classifying packets, as well as to have queues and mechanisms for their maintenance.

Traffic processing mechanism

The traffic processing mechanism includes:

802.1p;
Differentiated PER-HOP-BEHAVIORS services (DiffServ PHB);
Integrated services (intserv);
ATM et al.

Most local networks are based on IEEE 802 technology including Ethernet, token-ring et al. 802.1p is a traffic processing mechanism to support QoS in such networks.

802.1p defines the field (level 2 in network model OSI) in the header of the package 802, which can carry one of the eight priority values. As a rule, hosts or routers, sending traffic to local network, Mark each sent package, assigning it a certain value of priority. It is assumed that network devices, such as switches, bridges and hubs, process packets accordingly, using queue mechanisms. The scope of 802.1p is limited to the local network (LAN). As soon as the packet crosses the local network (via 3 OSI level), 802.1p priority is removed.

DiffServ is a level 3 mechanism. It defines the field in the level 3 IP packet header called DiffServ CodePoint (DSCP).

INTSERV is a whole range of services that defines a guaranteed service and service manager. The guaranteed service promises to carry some volume of traffic with measurable and limited delay. The service manager agrees to carry some volume of traffic with the "advent of light load." These are measurable services in the sense that they are defined to provide measurable QoS to a certain number traffic.

Since ATM technology fragments packets in relatively small cells, it can offer a very low delay time. If you need to transfer the packet Urgently, the ATM interface can always be released for transmission for the time you want to transmit one cell.

QoS has many more different complex mechanisms that ensure the work of this technology. We note only one important moment: In order for QoS to earn, support this technology and the corresponding setting throughout the transmission from the starting point to the final one.