KR100604197B1 - Apparatus and method for controling traffic in router - Google Patents

Abstract

The present invention relates to a method for controlling traffic in a network layer device, and upon receiving a packet, obtaining a destination IP address of the packet, and checking whether a network path corresponding to the destination IP address exists in a utilization database. And, if the network path exists, updating the utilization rate of the network path, comparing the current utilization rate of the network path with a predetermined threshold value, and determining whether the network path is congested. In the case of routing the packet to the bypass path, if not in the congestion state, calculating the difference between the current utilization rate and the threshold, and checking whether congestion is expected based on the difference, and the congestion expected If not, route the packet to the best path based on the routing table. The process of floating and, when the congestion is expected, including the step of routing the packet to alternate path. As such, the present invention can grasp the network state by using utilization information on each network path, and thus there is an advantage of predicting and preventing traffic congestion in advance.

Router, traffic, IP address, traceroute, network path, web data

Description

Apparatus and method for controlling traffic in a router {APPARATUS AND METHOD FOR CONTROLING TRAFFIC IN ROUTER}             

1 is a diagram illustrating a configuration of a router according to the present invention.

2 is a diagram for explaining network path grouping according to the present invention;

3 is a diagram illustrating a procedure for controlling traffic according to a network path state in a router according to an embodiment of the present invention.

4 is a diagram illustrating a procedure for mapping a specific destination to a representative path in a router according to an embodiment of the present invention.

5 is a graph showing the frequency of use of the representative path for each time zone.

The present invention relates to an apparatus and method for controlling traffic in a router, and more particularly, to an apparatus and method for controlling traffic using a network path.

The fastest developing field since the development of the Internet network is the web (WEB) field. Since web documents can accommodate all areas from text to multimedia, they can be rapidly developed due to the development of various contents. The type and number of web documents present on the Internet are immense. Even search engines for searching them have been developed and are being serviced. Therefore, what you need to use a huge web document is how to classify it and how to use it.

The most common use of web document classification is in search service. Web document classification is used to analyze usage by content and to identify user patterns or trends and apply them to other fields. The other field is the web cache server field, which similarly stores frequently used web documents through contents usage rates and provides them to users.

Currently, general web document classification is mainly based on contents included in web documents. The classification is classified into content type, scope, inquiry layer, and application scope. In other words, the classification for the purpose of the web document itself is not discussed. Although web data accounts for a large portion of the Internet data, the application field through web document classification is currently occupying most of the data. In other words, even though web data has a large proportion of data flowing through the network, it cannot be analyzed and used.

Since classification based on web content according to the prior art cannot be understood or judged at a network layer such as a router or a switch, it is unreasonable to control network traffic using web document classification. Therefore, there is an urgent need for a method of classifying web data so that it can be used in a layer below the application layer and controlling Internet traffic (web data) using the classification.

Accordingly, an object of the present invention is to provide an apparatus and method for classifying and analyzing web data in a lower layer rather than an application layer in an internet network, and controlling internet traffic using the analysis information.

Another object of the present invention is to provide an apparatus and method for analyzing a utilization rate according to a network path in a network layer device such as a router or a switch and controlling traffic using the analysis information.

It is still another object of the present invention to provide an apparatus and method for analyzing a utilization rate according to a web data path in a network layer device such as a router or a switch, and controlling traffic using the analysis information.

According to the present invention for achieving the above object, there is provided a method for controlling traffic in a network layer equipment, the process of acquiring a destination IP address of the packet when the packet is received, the network path corresponding to the destination IP address utilization rate Checking whether the network path exists in a database; if the network path exists, updating the utilization rate of the network path; and determining whether the network path is congested by comparing a current threshold value with a predetermined threshold value. And, in the congested state, routing the packet to a bypass path, and in the non-congested state, calculating a difference between the current utilization rate and the threshold value and checking whether congestion is expected based on the difference. And if the congestion is not expected, route the packet to a routing table. Routing to the optimal path based on the; and routing the packet to the bypass path if the congestion is expected.

Preferably, when the network path corresponding to the destination IP address does not exist in the utilization database, storing the destination IP address in a specific area of the memory; Reading a destination IP address, performing a trace route with the destination IP address, creating a network path, checking whether the created network path exists in the utilization database, and if present, the destination And storing the IP address in the utilization database corresponding to the confirmed network path, and if not present, adding the created network path to the utilization database as one record.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In describing the present invention, if it is determined that the detailed description of the related known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted.

The web data classification proposed in the present invention is a classification according to a network path through which a packet flows. The Internet network consists of routers configured by hop by hip. Requesting web data and sending and receiving a response to the request through the router from the current user location to the server where the web document is stored. Based on this configuration, it is to identify and classify the network path to the server where each web data exists. Once classified, traffic information can be collected accordingly. In other words, it is possible to obtain information on which path the web data is used most and how much is used. Next, traffic is controlled by collecting information about web data paths, analyzing patterns and utilization rates.

1 shows a configuration of a router according to the present invention. In the following, a distributed router, in which many functions are distributed to line cards, will be described as an example.

As shown, the router includes a plurality of line card modules (LCMs) 111-116, a switch 130, and a processor 140. In this case, the processor 140 includes a routing table 141 and a utilization database (DB) 142, and each line card module includes a forwarding table (FT).

Referring to FIG. 1, the processor 140 includes a memory and controls the overall operation of the router. Basically, it processes the routing protocol that manages the routing table 141 in which the best route is stored. Here, the processor 140 immediately updates a changed route (path) by executing a routing protocol such as Routing Information Protocol (RIP), Open Shortest Path First (OSPF), or Border Gateway Protocol (BGP). According to the present invention, the processor 140 manages a utilization database 142 in which the utilization rate for each network path (web data path) is stored, and based on the utilization database 142, Function to control the flow. The processor and the memory may be physically configured on one chip or may be configured as separate chips.

The line card modules (LCMs) 111-116 each have a port connected to a network or a host, and mainly serve as an external interface function. Since the interface port of the line card module provides a local area network (LAN) and wide area network (WAN) interface, another router or host may be physically connected, and the workstation server may be connected in a LAN form. The following description will be described assuming that other adjacent routers are connected.

The forwarding tables 121-126 provided in the line card modules 111-116 are provided to provide an efficient lookup in each line card module. The routing information downloaded from the routing table 141 is provided. Save it. That is, the packet received by the line card module is forwarded to another line card module based on the forwarding table. Packets forwarded to other line card modules are then routed to other adjacent routers.

On the other hand, the switch 130 is responsible for connecting the processor 140 and the line card modules (111-116) to enable mutual communication. The switch 130 switches packets exchanged between the processor 140 and the line card modules 111-116 and relays the packets.

Although not shown, for example, the processor 140 and the line card modules 111-116 may communicate through inter-processor communication (IPC).

Looking at the operation based on the configuration of Figure 1 as follows.

For example, when a packet is received by the line card module 111, the line card module 111 reports the destination IP address of the packet to the processor 140 through IPC. The processor 140 then searches the utilization database 142 to check if a network path exists for the destination IP address. If the network path exists, the current utilization rate of the network path is checked, and if the current state is normal, the line card module 111 is instructed to route the packet to the optimal path. Then, the line card module 111 routes the packet to the optimal path based on the forwarding table. If the current state is congested or congestion is expected, the processor 140 instructs the line card module 111 to route the packet to another path. Then, the line card module 111 routes the packet to a default or randomly selected path. As described above, the present invention is characterized by controlling the traffic according to the current network path state.

Hereinafter, the overall operation according to the present invention will be described.

First, the present invention collects information about web data transmitted through the Internet network. Generally, web data uses port number 80. Therefore, the source and destination IP addresses of the packet using port number 80 are collected.

Next, we use a tool called traceroute on the collected IP addresses to determine the network path to the source and destination IP addresses. In the following description, it is assumed that the route is identified only for the destination IP address.

The trace route uses a time to live (TTL) of the IP datagram. The source router running the traceroute sends multiple datagrams to specific destinations. In the first datagram, the time-to-live field is set to '1'. This causes the datagram to be destroyed at the first router on the path. The first router sends a 'datagram extinction' message back to the source router. This message contains in the header the address of the router sending it, from which it identifies the first router on the path. Thereafter, the source router transmits the datagram having the time-to-live field increased by '1' back to a specific destination. This datagram is destroyed at the next router (second router) on the path, and the second router sends a 'datagram destroyed' message back to the source router. In this way, the addresses of hops (routers) present on a particular URL path are obtained.

Next, the route for the obtained destination IP address is mapped to the representative route. A representative path is set by grouping similar paths among the paths to reach the destination. For example, if the route to reach naver.com is a-b-c-d and the route to daum.net is a-b-c-e, the routes to these two destinations are mapped to the representative route a-b-c. As such, the paths passing through the common path (or the representative path) are managed as a group.

2 illustrates network path grouping. In the drawing, paths (or destinations) belonging to group A are mapped to representative path 1, and paths belonging to group B are mapped to representative path 2. As such, the reason for grouping paths is that problems such as load may occur when traffic control is performed with paths to all destinations.

Next, grasp the utilization rate (frequency of use) based on the representative path. The utilization rate is then used for traffic flow control. If congestion is expected for a particular representative path from the utilization, then an incoming packet is sent to a path other than the optimal path managed in the routing table.

As such, the present invention can grasp the flow for a specific network path. Most domestic servers use IDC (Internet Data Center), so they can understand the flow of specific IDC paths. Hereinafter, specific embodiments will be described with reference to the accompanying drawings.

3 illustrates a procedure for controlling traffic according to a network path state in a router according to an embodiment of the present invention.

Referring to FIG. 3, the processor 140 first checks whether a packet is received by the line card modules 111-116 in step 301. If the packet reception is detected, the router checks the port number of the packet header in step 303. In step 305, the router checks whether the received packet is web data based on the port number. If it is not the web data, the processor 140 proceeds to step 329 to perform the corresponding function.

If the received packet is web data, the processor 140 proceeds to step 307 to obtain a destination IP address of the received packet. In operation 309, the processor 140 searches the utilization database 142, and in operation 311, the processor 140 checks whether a representative path corresponding to the destination IP address exists.

If the representative path corresponding to the destination IP address does not exist, the processor 140 proceeds to step 325 and stores the destination IP address in a specific area of the memory and proceeds to step 327. If there is a representative path corresponding to the destination IP address, the processor 140 updates the utilization database in step 313. For example, the utilization database is configured as shown in Table 1 below.

Representative Path Identifier Route information (router's address string) Destination addresses via representative path Utilization 00-08 08-13 o'clock 13-18 o'clock 18-00 o'clock First representative path (First Address, Second Address, ..., X Address) ..... 20% 30% 30% 20% 2nd Representative Path (First Address, Second Address, ..., Y Address) ...... 10% 40% 20% 10% ... ... ..... ... ... ... Nth representative path (1st address, 2nd address, ..., Z-address) ..... 20% 20% 20% 20%

As shown in Table 1, the utilization database stores a representative path identifier, address strings of routers belonging to the representative path, destination addresses via the representative path, and time-of-use utilization information.

If the destination address of the received packet is among the destination addresses of the utilization database, the processor 140 updates the utilization corresponding to the representative path. If the current time is 15:00, the utilization rate corresponding to the specific time zone (13:00-18:00) is updated. In general, the utilization rate is calculated by dividing the total number of packets in a specific time zone by the number of packets in the representative path. Therefore, the result value is output as 13% at 1-18% utilization of the first representative path, 20% of the second representative path.

The utilization database as described above may be initialized at a predetermined period (eg, daily unit). Also, even if initialized, data for a certain period of time (or average data for three days) is kept. This historical data can be used to predict future network conditions.

After updating the utilization database as described above, the processor 140 proceeds to step 315 to determine the current state of the representative path for the destination IP address. That is, the current utilization rate (called a first utilization rate) of the representative path is read from the utilization database.

In operation 317, the processor 140 compares the first utilization rate with a predetermined threshold (eg, 70%) to check whether the processor is congested. That is, it is checked whether the first utilization rate is greater than the threshold value. If determined to be congested, the processor 140 proceeds to step 323 and transmits the received packet to a path other than the representative path. If the routing table 141 stores not only the optimal routing path but also the lane routing path, the received packet transmits the lane routing path, but otherwise, it transmits through the default or randomly determined other path. do.

If it is determined that there is no congestion, the processor 140 reads a utilization rate (called a second utilization rate) corresponding to the current time of the representative path from previously stored historical data in step 319, and reads the read second utilization rate. Determine if congestion is expected from As another example, the difference between the threshold value and the first utilization rate is calculated, and if the difference is smaller than a preset value, it may be determined that the congestion expected state. If it is determined that the congestion is expected, the processor 140 proceeds to step 323 and transmits the received packet to another path.

If it is determined that the congestion is not expected, the processor 140 proceeds to step 327 and transmits the received packet to the optimal path according to the routing table 327.

4 illustrates a procedure for mapping a specific destination to a representative path in a router according to an exemplary embodiment of the present invention.

Referring to FIG. 4, first, the processor 140 checks whether a path creation event occurs. The routing event is programmed to occur at off-load times. When the route creation event occurs, the processor 140 reads a destination IP address stored in a specific area of the memory in step 403. The destination IP address is stored in step 325 of FIG.

After reading the destination IP address, the processor 140 performs a traceroute to the destination IP address in step 405. The traceroute may be performed using an internet control message protocol (ICMP). As described above, the traceroute is a method of confirming the existence of a router on a path by sending an ICMP echo message from each router and confirming a response thereto.

As such, after confirming the existence of the routers existing on the path, the processor 140 creates a network path (a type of address string of routers) with the addresses of the identified routers in step 407. In operation 409, the processor 140 searches the utilization database 142 using the created network path, and in step 411, the processor 140 determines whether there is a representative path that matches (eg, 70%) the network path. do.

If the representative path does not match to some extent, the processor 140 adds the network path to the utilization database 141 as the representative path in step 415 and proceeds to step 417. If there is a representative path that matches to some extent, the processor 140 proceeds to step 413 and stores the destination IP address in response to the identified representative path and proceeds to step 417.

In operation 417, the processor 140 checks whether another destination IP address is stored in a specific area of the memory. If there is another destination IP address, the processor 140 returns to step 403 to perform the following steps again. If there is no more destination IP address, the processor 140 terminates the algorithm.

5 is a graph showing the frequency of use of the representative path for each time zone. In the figure, the horizontal axis represents the network path (representative path) and the vertical axis represents the required number (frequency of use). As shown in the figure, when information such as utilization rate and frequency of use for the network path is grasped, the characteristics of network traffic can be predicted to some extent. In other words, if the load information for a particular path can be expected, congestion can be prevented or avoided based on this, and a bypass path can be selected at the time of congestion. This enables the hop by hop method to know and cope with the current network status by using frequency and current utilization information of each path in a routing environment where the overall network status is unknown.

Meanwhile, in the detailed description of the present invention, specific embodiments have been described. However, various modifications may be made without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined not only by the scope of the following claims, but also by those equivalent to the scope of the claims.

As described above, the analysis of the web data for each network path according to the present invention can be easily used in efficient network operation or additional network configuration in the future. Above all, the current network utilization can be used to determine the network status, which can be used to predict and prevent traffic congestion.

Claims (10)

An apparatus for controlling traffic in network layer equipment, A database for managing each network path including identification information, address strings of network layer devices existing in the network path, destination IP addresses and utilization rate information via the network path, and And upon receiving the packet, accessing the database to update the utilization rate of the network path corresponding to the destination IP address of the packet and routing the packet to the optimal path or the bypass path based on the utilization rate of the network path. Characterized in that the device. delete The method of claim 1, And said network layer equipment is a router or a switch. The method of claim 1, And the processor determines a congestion state by comparing a current utilization rate of the network path with a preset threshold, and routes the packet to the bypass path in the congestion state. In the method for controlling traffic in the network layer equipment, Initializing a utilization database managing a utilization rate for a network path corresponding to a column of network layer devices at a predetermined time period; When the packet is received, obtaining a destination IP address of the packet; Checking whether a network path corresponding to the destination IP address exists in the utilization database; Updating the utilization rate of the network path when the network path exists; Determining whether the network path is congested by comparing a current utilization rate of the network path with a predetermined threshold value; In the congested state, routing the packet to a bypass path; If not in the congestion state, calculating a difference between the current utilization rate and the threshold, and checking whether congestion is expected based on the difference; If the congestion is not expected, routing the packet to an optimal path based on a routing table; If the congestion is expected, routing the packet to a bypass path. The method of claim 5, When the network path corresponding to the destination IP address does not exist in the utilization database, storing the destination IP address in a specific area of the memory; Reading the destination IP address from a specific area of the memory when a route creation event occurs; Creating a network route by performing a trace route with the destination IP address; Checking whether the created network path exists in the utilization database; If present, storing the destination IP address in the utilization database corresponding to the identified network path; If it does not exist, the method further comprising the step of adding the created network path to the utilization database as one record. The method of claim 5, The network layer equipment is a router or a switch. delete The method of claim 5, wherein the utilization database, And storing identification information, address strings of network layer devices existing in the network path, destination IP addresses through the network path, and utilization rate information for a specific network path. The method of claim 9, Addresses of network layer devices present in the network path are obtained using a traceroute of an internet control message protocol (ICMP).

Images