KR102435334B1 - Adaptive Bidirectional Forwarding Detection protocol and equipment for maximizing service availability in network system - Google Patents

Abstract

A method for transmitting a control message by a first device of a BFD system according to an embodiment of the present invention includes: sequentially transmitting a set of echo messages having different transmission periods to a second device; optimizing the BFD protocol, including determining a transmission period of a control message based on a loopback result of the set of transmitted echo messages and transmitting the control message to the second device according to the determined transmission period It can increase the availability of users of the network.

Description

Adaptive Bidirectional Forwarding Detection protocol and equipment for maximizing service availability in network system

The present invention relates to a method and apparatus for recovering a failure within the shortest possible time when a failure occurs in a network system so that a user can be provided with a seamless service. Specifically, the service availability of the network system is improved. It relates to an adaptive BFD (Bidirectional Forwarding Detection) protocol and apparatus for maximally guaranteeing.

In general, a failure in a network system can be divided into a node failure and a network failure. Node failure is detected by considering the internal network resources and environment of a specific node system, and specifically, the average load and usage rate of each device or the bandwidth of the link in which the BFD operates. It is detected considering the external congestion situation.

One of the most important issues in operating a network system is to detect and resolve failures as quickly as possible when failures occur due to various causes. Service Level Agreement) management is becoming more and more important.

In recent years, in relation to service recovery within a short time in the event of a failure, various technologies have been proposed that provide a seamless service, that is, recover from a failure in such a short time that the network user does not recognize the fact that the failure has occurred. have. Among them, the BFD Protocol standardized by the Internet Engineering Task Force (IETF) is attracting attention as a failure detection protocol. The node failure and network failure can be quickly detected by the BFD Protocol.

The present invention proposes a method for increasing the availability of users of a network by optimizing the BFD protocol.

A method for transmitting a control message by a first device of a BFD system according to an embodiment of the present invention includes: sequentially transmitting a set of echo messages having different transmission periods to a second device; and determining a transmission period of a control message based on a loopback result of the set of transmitted echo messages, and transmitting the control message to the second device according to the determined transmission period.

The BFD device according to an embodiment of the present invention sequentially transmits a set of echo messages having different transmission periods to a second device, and transmits the control message to the second device according to the transmission period. and a communication unit for transmitting to the , and a control unit for determining the transmission period of the control message based on a loopback result of the set of the transmitted echo messages.

During BFD operation, it is possible to minimize traffic loss that occurs when a failure occurs by setting a timer that dynamically reflects network congestion that changes in real time. Accordingly, the BFD itself continuously and repeatedly adapts to the network environment, thereby increasing the service availability experienced by users. It is possible to determine an appropriate control packet transmission timer (tx timer) according to a network environment while reducing effort or trial and error through complicated traffic engineering or trial and error methods by a network administrator.

1 shows a topology of a mobile communication network operating BFD.
2 shows a general scenario in which BFD is operated.
3 is a BFD Control Protocol Data Unit (PDU) format 301 and a BFD Echo PDU format 304 defined in RFC 5880.
4 is a process for a BFD message flow defined in RFC 5880 (Message flow).
5a, 5b, 5c and 5d are operation procedures of each module when operating the BFD protocol.
6 shows the overall operation flow of the BFD to which the adaptive timer is applied.
7 shows the overall message flow of the adaptive BFD protocol.
8 is a graph showing the effect of a timer value on transmission loss, failure recovery time, and service availability.
9 shows adjustment of a timer value of a BFD echo packet in a timer adaptation process.
10 illustrates an example of performing a timer adaptation operation based on throughput.
11 shows an embodiment of the modules included in the system of the present invention.
12a, 12b, 12c and 12d are flow charts showing the internal operation process of the BFD Protocol according to the present invention.
13 shows a method for a device of a BFD system to transmit a control message.
14 shows the configuration of a BFD device.

In this specification and claims, “comprising” does not mean excluding other elements or acts. In this specification and claims, singular nouns may include plural nouns unless specifically stated otherwise. For example, "packet" may refer to one packet or may include two or more packets. Also, "message" may refer to one message or may include two or more messages. In the present specification and claims, the suffix “part” for a component is given or mixed in consideration of only the ease of writing the specification, and does not have a meaning or role distinct from each other by itself. In this specification and claims, "first, second, and third" are used to distinguish similar elements, and are not necessarily used to describe sequentially or chronologically.

In this specification and claims, "packet", "traffic", and "message" may be used interchangeably. In this specification and claims, "packet loss", "traffic loss", and "transmission loss" may be used interchangeably. In this specification and claims, "timer", "interval", and "cycle or period" may be used interchangeably. When a packet or message has a transmission period, the packet or message includes two or more packets or messages. In this specification and claims, "link" and "session" may be used interchangeably. In the present specification and claims, each subject that transmits or receives a “packet” may be expressed as a “peer”, “device”, “system”, and the like.

Hereinafter, various embodiments of the present invention will be described with reference to the accompanying drawings. In the following description, only the parts necessary for understanding the operation according to the embodiment of the present invention are described, and the description of other parts is simplified or omitted so as not to obscure the gist of the present invention. Here, the features of the present invention are not limited to the above-described examples, and may include a change in the shape of each configuration described below or even additional functions. In the drawings, the sizes of some elements may be enlarged for illustration, and are not drawn to scale.

1 shows a topology of a mobile communication network operating BFD.

When a failure occurs in the link between the Switch-1 110 and the Router D 130 (101), the Router B 120 and the Router D 130 use the BFD Protocol to the Router B ( 120) and the Router D 130 detects a session down (session down) (102). Then, Router B 120 and Router D 130 deliver a failure notification to the client for each failure (103), delete neighbor information of the client for the down session (104), and finally Traffic repairing to a secondary path is performed (105).

2 shows a general scenario in which BFD is operated.

For the operation of the BFD Protocol, session establishment, timer negotiation, and fault detection are performed between R1 (210) and R2 (220) (201). When traffic loss is detected during traffic forwarding 202 in the primary path, traffic forwarding 203 to the secondary path is performed through a failure repairing process.

3 is a BFD Control Protocol Data Unit (PDU) format 301 and a BFD Echo PDU format 304 defined in RFC 5880.

The BFD Control PDU 301 includes a basic field (Version, Discriminator, etc.), an upper area such as a Desired Min TX Interval 302 and a Required Min RX Interval 303, and a UDP field (Destination Port: 3784). The BFD Echo PDU 304 includes a basic field, a Required Min Echo RX Interval 305, an upper field such as a Required Min RX Interval 303, and a UDP field (Destination Port: 3785). The BFD Control PDU 301 has a uni-directional characteristic and the BFD Echo PDU 304 has a bi-directional characteristic.

4 is a process for a BFD message flow defined in RFC 5880 (Message flow).

4 shows a message flow for performing Session Establishment, Timer Negotiation, and Fault Detection between a local system and a remote system. In the Session Establishment 401 step, a BFD session is created using a three-way handshake method of the BFD Control packet. As shown in FIG. 4 , as the BFD Control packet is transmitted, the session state sequentially changes to Down, Init, and Up.

In the Timer Negotiation 402 step, the timer period of the BFD session is determined using a poll sequence. The Poll sequence includes transmission (reception) of a BFD Control packet with Poll bit and reception (transmission) of a BFD Control packet with Final bit. . In FIG. 4 , the remote system transmits a Desired Min Tx Interval (100 ms) and a Required Min Rx Interval (100 ms) to the Local System through a BFD Control packet with Poll bit. The Tx interval of the remote system corresponds to the Rx interval of the local system, and the Rx interval of the remote system corresponds to the Tx interval of the local system. Tx interval and Rx interval are values corresponding to each session connecting two systems to each other.

The Local System sets the Desired Min Tx Interval (100ms) and Required Min Rx Interval (100ms) of the Remote system obtained through the BFD Control packet with Poll bit to the Required Min Rx Interval (150ms) and Desired Min Tx Interval (150ms) of the Local System. ) and the Negotiated Min Rx Interval and Negotiated Min Tx Interval of the local system can be determined. The Negotiated Min Rx Interval may be, for example, 150 ms, which is a larger value among a Desired Min Tx Interval (100 ms) of a remote system and a Required Min Rx Interval (150 ms) of a local system. The Negotiated Min Tx Interval may be, for example, 150 ms, which is a larger value among Required Min Rx Interval (100 ms) of the remote system and Desired Min Tx Interval (150 ms) of the local system. When the Local System transmits the Negotiated Min Tx Interval (150 ms) to the Remote System through the BFD Control packet with Final bit, the Remote System may determine the Negotiated Min Tx Interval of the Local System as the Negotiated Min Rx Interval of the Remote System. When the Local System transmits the Negotiated Min Rx Interval (150 ms) to the Remote System through the BFD Control packet with Final bit, the Remote System may determine the Negotiated Min Rx Interval of the Local System as the Negotiated Min Tx Interval of the Remote System. Unlike FIG. 4 , when the Local System transmits the BFD Control packet with Poll bit to the Remote System, the Remote System determines the Negotiated Min Tx Interval and the Negotiated Min Rx Interval, and transmits the BFD Control packet with Final bit to the Local System.

In the Fault Detection (403) step, fault detection is performed by periodically checking the status of the BFD session based on the determined interval. In FIG. 4 , the local system and the remote system transmit the BFD Control packet using the Negotiated Min Tx Interval determined in the Timer Negotiation step. In FIG. 4 , the Local system and the Remote system determine whether the BFD Control packet reception interval and the Negotiated Min Rx Interval determined in the timer negotiation step have a difference of less than or equal to a reference value, whether or not BFD Control packet loss occurs to judge

5a, 5b, 5c and 5d are operation procedures of each module when operating the BFD protocol.

5A shows the operation of the BFD-Session Management module 501 . The BFD-Session Management module 501 manages session parameters including creation/deletion/update and timer for the session of the BFD. For example, the BFD-Session Management module 501 stores the Session Configuration (505), and stores the Session Timer Configuration (506).

5B shows the operation of the BFD-Tx module 502 . The BFD-Tx module 502 transmits a BFD control packet to the peer system based on the BFD session configuration information. For example, the BFD-Tx module 502 reads the session setting information and the session timer setting information stored by the BFD-Session Management module 501 (507), and creates a Session based on the read information (508), The BFD packet is transmitted to the peer system through the created Session (509). The transmission of the packet is terminated when there is no packet to be transmitted or when the session state is not UP ( 510 ).

5C shows the operation of the BFD-Rx module 503 . The BFD-Rx module 503 checks and changes the session state according to the received packet type. For example, the BFD-Rx module 503 receives the BFD packet ( 511 ), and classifies the type of the received BFD packet ( 512 ). The BFD-Rx module 503 checks the session state (513) to determine whether the session state is Down (514). Under certain conditions, the BFD-Rx module 503 changes the session state to Up if the session state is Down (515), and clears the Detection Timer if the session state is not Down (516).

5D shows the operation of the BFD-Event Management module 504 . The BFD-Event Management module 504 performs fault detection by monitoring the timer. For example, the BFD-Event Management module 504 checks the timer (517) and when it is determined that the timer has expired (518) changes the session state from Up to Down (519).

The BFD protocol detects system and network failures through transmission and reception of BFD control messages. The BFD timer, which means the transmission period of the BFD control message, may have a default value set based on the network operator's experience, network propagation delay, average load and usage rate of each device, or performance of each device. Once the BFD timer is set, it is difficult to dynamically reflect real-time traffic conditions within the network.

When the tx period of the BFD packet is short (when the BFD timer has a small value), a failure can be detected quickly and service availability can be maximized. However, since BFD packet transmission also occupies resources, an increase in the total amount of transmission increases the load on the network. That is, the BFD packet applied for fast failure detection may rather affect the existing network flow, so that the BFD may perform an erroneous failure detection operation.

Conversely, when the tx period of the BFD packet is long (when the BFD timer has a large value), the load on the system and network can be reduced by reducing the transmission amount of the BFD packet, but the time it takes for the BFD to detect a failure and recover will increase

After the BFD session timer negotiation is completed, if communication with a neighbor fails or session flapping occurs at a specific point in time, the BFD timer value should be changed to a larger value from the viewpoint of reducing the effect caused by the BFD operation.

The present invention discloses a method of applying a new BFD protocol operation by acquiring a timer in consideration of current network resources and traffic conditions by performing a timer adaptation operation before a network failure occurs. In other words, we propose a method of operating BFD with the timer most suitable for the current state by reflecting the continuously changing network congestion information in real time, rather than operating based on a one-time acquired timer.

The present invention is not a reactive method to respond after a network failure occurs, but proactively sets an optimized timer in advance.

6 shows the overall operation flow of the BFD to which the adaptive timer is applied.

For operation of the BFD protocol, session establishment, timer adaptation, timer negotiation, and fault detection are performed between R1 610 and R2 620 (601). Comparing step 601 with step 201 of FIG. 2 , it can be seen that a timer adaptation process is added. The timer adaptation process will be described later.

After the message flow of step 601 is completed, the primary traffic transmitted between R1 and R4 is forwarded on the session in which the BFD Protocol is operating (602).

When traffic loss occurs for the primary traffic transmitted between R1 and R4, the BFD Protocol performs failure notification and repairing (603).

Secondary traffic transmitted between R1 and R4 after being repaired by the BFD Protocol is changed to go through R3 instead of R2 (604).

7 shows the overall message flow of the adaptive BFD protocol.

Each peer (system) of the BFD creates a session using the 3-handshake method of the BFD control packet and performs a session establishment process of changing the state of the session to 'down → init → up' (701).

Each peer of BFD applies the BFD echo packet loop-backed from the remote system to the network to determine the optimal timer before performing timer negotiation, and is the fastest way to detect faults in the current state depending on whether loss occurs or not. A timer adaptation process for finding a sweet spot' (705) point is performed (702).

The timer adaptation process may include determining whether loss has occurred in the EFD echo packet returned by loop-back when the BFD echo packet is transmitted to the remote system using the echo timer having a predetermined value. When the EFD echo packet has a value of echo timer T1, it means that two or more EFD echo messages are transmitted at a time interval of T1. In case of loss, it is possible to retransmit the BFD echo packet by greatly changing the value of the echo timer. If the loss does not occur, the BFD echo packet may be retransmitted by changing the value of the echo timer small. In this way, if a preset condition is met while repeating transmission of the BFD echo packet while changing the size of the echo timer, the BFD echo packet when the preset condition is satisfied can be determined as a sweet spot, and the sweet spot The value of the echo timer of the BFD echo packet determined as the spot can be determined as the optimal timer. 7 shows that the value (350ms) of the echo timer of session 5 transmitted for the fifth time is determined as the optimal timer (705). The Optimum timer is used as the desired Min Tx interval of the local system in the timer negotiation process.

Each peer of the BFD performs a timer negotiation process using the BFD poll sequence procedure to inform other peers of the optimal timer determined through the timer adaptation process (703). In the timer negotiation process, as described with reference to FIG. 4 , a BFD Control packet with Poll bit and a BFD Control packet with Final bit are transmitted and received.

Each peer of the BFD transmits the BFD Control packet to the peer that has performed the timer negotiation process for a certain period of time or until there is a user's input with the optimal timer period (704). When traffic loss occurs, as a result of being repaired by the BFD Protocol, the flow of traffic is changed in the direction of the secondary traffic path.

8 is a graph showing the effect of a timer value on transmission loss, failure recovery time, and service availability.

Tx Timer means a period in which the local system transmits a BFD Control packet (BFD packet) for fault detection to the remote system. A small value of Tx Timer means an increase in system and network load by the BFD Control packet. When the period at which the BFD Control packet is transmitted is smaller than a predetermined value, a transmission loss 801 due to the BFD Control packet occurs. The Valid Tx Timer 804 corresponding to the transmission period of the BFD Control packet in which the transmission loss starts to occur varies depending on the current network resource and traffic conditions. The current Tx timer is preferably larger than the Valid Tx Timer 804 . On the other hand, in terms of a fail-over time 802 , the current Tx timer is preferably smaller than the valid Tx Timer 804 . . This is because the smaller the Tx timer, the faster the failure can be detected. 8 shows that the current Tx timer is preferably smaller than the valid Tx Timer 804 in terms of service availability 803 .

9 shows adjustment of a timer value of a BFD echo packet in a timer adaptation process.

A binary search may be used for timer adaptation, which is shown in FIG. 9 . According to binary search, if loss occurs, the timer value is increased and BFD echo packet is transmitted. If loss does not occur, the timer value is decreased and BFD echo packet is transmitted. At t1, a loss occurred according to the loopback result of transmitting the echo packet with the Tx Timer set to 50 ms. Therefore, the Echo packet was transmitted with the Tx Timer set to 100 ms (double of 50 ms) at t2. At t2, the loss occurred according to the loopback result of transmitting the echo packet with the Tx Timer set to 100 ms. Therefore, the Echo packet was transmitted with the Tx Timer set to 200 ms (double 100 ms) at t3. Since no loss occurred according to the loopback result of transmitting the echo packet with the Tx Timer set to 200 ms at t3, the Tx timer is set to 150 ms at t4 (the Tx timer of the most recent echo packet that has lost and no loss occurs). Echo packet is transmitted with average of Tx timer of the last echo packet that has not been reached: (100+200)/2 ). At t4, the loss occurred according to the loopback result of transmitting the echo packet with the Tx Timer set to 150 ms. Therefore, the Tx Timer was set to 175 ms at t5 (Tx timer of the most recent echo packet with loss and Tx of the most recent echo packet with no loss) Echo packet was transmitted with timer average: (150+200)/2). At t5, the Tx Timer was set to 175ms and no loss occurred according to the loopback result of transmitting the echo packet. Therefore, the Tx Timer was set to 162.5ms at t6 (Tx timer of the last echo packet with loss and the last echo without loss) Echo packet was transmitted with average of packet Tx timer: (150+175)/2). At t6, the Tx Timer was set to 162.5ms and no loss occurred according to the loopback result of transmitting the echo packet, so at t7, the Tx Timer was set to 156.25ms (Tx timer of the last echo packet with loss and the most recent Echo packet with no loss). Echo packet was transmitted with average of Tx timer of echo packet: (150+162.5)/2). In FIG. 9 , it is determined that the Tx Timer of the echo packet transmitted at t7 satisfies the adaptation end condition, and the Tx Timer (156.25 ms) of the echo packet transmitted at the time the adaptation end condition is satisfied (t7) is optimal. It indicates that it is determined by the Optimum Tx timer.

In FIG. 9 , the Tx Timer of the echo packet is doubled for every echo packet transmission until the first lossless echo packet is transmitted. This is an exponential increase and can also be implemented by increasing by a number (for example, 1.5 times) that is not doubled. Also, it may be implemented as an increase by a certain value (eg, 10 ms) rather than an exponential increase.

In FIG. 9, after transmitting an echo packet with a loss, the Tx timer of the echo packet is determined as the average value of the Tx timer of the most recent echo packet with loss and the Tx timer of the most recent echo packet with no loss for every echo packet transmission. . Referring to FIG. 9 , it can be seen that the Tx timer of the recent echo packet in which loss has occurred has a smaller value than the Tx timer of the recent echo packet in which loss has not occurred.

Unlike FIG. 9 , current link utilization may be reflected when determining the Tx timer of the echo packet. Link utilization has a value between 0 and 1, and the relationship between the maximum capacity (Capacity) of each link and link utilization (util) is shown in Equation 1 below.

If util is greater than a certain value (eg 75%), the increase rate can be decreased and the decrease rate can be increased when obtaining the Tx timer of the next echo packet. For example, when determining the Tx timer at t2 in FIG. 9 , it may be set to 80 ms corresponding to 1.6 times, not double, 50 ms. Alternatively, when calculating the Tx timer at t6, the Tx timer may be set to a value smaller than 162.5 ms by giving a large weight to the Tx timer 150 of the last lost echo packet. If util is less than a certain value (eg 50%), the increase rate can be increased and the decrease rate can be decreased when obtaining the Tx timer of the next echo packet. For example, when determining the Tx timer at t2 in FIG. 9 , it may be set to 125 ms, which is 2.5 times the 50 ms, not twice the 50 ms. Alternatively, when calculating the Tx timer at t6, the Tx timer (175ms) of the most recent echo packet with no loss may be given a large weight, and the Tx timer may be set to a value greater than 162.5ms. When util has a value within a certain range (for example, between 50 and 75%), the Tx timer of the next echo packet can be obtained by the method of FIG. 9 .

In the above example, link utilization is reflected by dividing into three sections, and it can also be implemented in a form of dividing into two sections or four or more sections. It can also be implemented in the form of directly reflecting link utilization in the calculation of the Tx timer without dividing the section.

The adaptation ending condition may be implemented in various ways. For example, the difference between the Tx timer (A) of the most recent echo packet with no loss and the Tx timer (B) of the most recent echo packet without loss, which is used to find the Tx Timer of the next echo packet, is the reference value (e.g. 10 ms) may correspond to the above condition. Alternatively, the above condition may correspond to the fact that the number of transmission times of the lost Echo packet is greater than the reference value (eg, 10 times). Alternatively, the above condition may correspond to the number of times of transmission of an echo packet in which no loss has occurred is greater than a reference value (eg, 5 times). Alternatively, the lapse of a certain time after transmission of the first echo packet may correspond to the above condition. Alternatively, the lapse of a certain time after transmission of the first echo packet in which no loss occurs may correspond to the above condition. The adaptation termination condition is not limited to the above-described examples, and a mixture of two or more of the above-described examples may correspond to the adaptation termination condition. When the adaptation termination condition is satisfied, the Tx Timer of the echo packet transmitted at the time when the adaptation termination condition is satisfied or the Tx Timer of the first transmitted echo packet after the condition is satisfied may be determined as the optimal Timer.

10 illustrates an example of performing timer adaptation operation based on throughput.

Throughput (TH) 1001 indicates the amount of packet (data) transmission, and may be mixed with bandwidth (BW). Since TH is proportional to the reciprocal of the Tx Timer 1002, TH can be used when an actual algorithm is implemented to determine the Tx Timer, and timer adaptation using TH is shown in FIG. 10 .

The device of the BFD transmits the BFD echo packet to another device by applying an initial TH value to determine the TH corresponding to the Tx Timer. The initial value may correspond to Max BandWidth of link 1003 . The timer corresponding to the Max BandWidth of link may correspond to the smallest cycle in which the BFD echo packet can be transmitted in the corresponding link. If the initial value of BandWidth is smaller than a predetermined value, a phenomenon may occur that the converging BandWidth according to timer adaptation becomes smaller than an actual available BandWidth. If the Max BandWidth of Link is set to the initial value, the probability of the above phenomenon occurring can be reduced. The BFD changes the timer value during timer adaptation according to the loss of the echo packet returned after loop-back. For timer adaptation, binary search method can be used as shown in the table below.

Initial TH value = Max BW of Link (1)
Initial T = (MAX + MIN)/2, TH/2 (2)
(MAX = TH, MIN = 0)
Initial Tx timer = 1/T
When packet loss occurs, T = (MAX + MIN)/2, (0+TH/2)/2 (3)
(MAX = TH/2, MIN = 0)
When no packet loss occurs, T = (MAX + MIN)/2, (TH+TH/2)/2 (4)
(MAX = TH, MIN = TH/2)

As can be seen from the equation shown in Table 1, timer adaptation is performed according to the presence or absence of Packet LOSS.

The timer adaptation procedure is terminated when an adaptation ending condition as shown in 1006 of FIG. 10 is satisfied, and the Tx timer value at the time of termination may be determined as the final timer value. The adaptation ending condition is not limited to the condition shown in FIG. 10 . For example, after transmitting the BFD echo packets of the initial value, the number of BFD echo packets having a TH value in which packet loss does not occur in the process of sequentially transmitting BFD echo packets having different TH (timer) values is equal to or greater than a predetermined number In this case, timer adaptation may be terminated. In this case, the timer of the last transmitted BFD echo packet may be determined as the final timer value.

11 shows an embodiment of the modules included in the system of the present invention.

Configurations related to the operation of the present invention include an Adaptation Timer configuration 1101 , a Session configuration 1102 , and a Session Timer configuration 1103 . BFD Session Management unit includes BFD Session Manager module(1104), Session Data module(1105), BFD Timer Manager module(1106), Timer Data module(1107), Timer Adapter module(1108) and BFD Fault Manager module(1114). include The Forwarding Session (Tx) unit includes a BFD PDU Generator (1109) and a BFD Packet Sender (1110). Forwarding Session (Rx) unit includes a BFD Packet Receiver (1112) and Session Event Manager (1113). The Peer unit of the Remote System includes a BFD Protocol Reply module 1111. 11 is an embodiment related to the operation of the present invention, and may be implemented differently.

12a, 12b, 12c, and 12d are flow charts showing the internal operation process of the BFD Protocol according to the present invention.

Comparing with FIGS. 5A to 5D, it can be seen that colored motions are added in FIGS. 12A to 12D. BFD-Session Management (1201) stores the session setting information that came down initially (1205), and stores adatation-related setting information in addition to storing the timer setting information of the session (1206) (1207). BFD-Tx (1202) Checks the adaptation flag information received from the BFD-Session Management (1201) (1208). When the adaptation flag is enabled, the BFD-Tx 1202 reads adaptation timer information (1215), generates an echo packet PDU (1216), and transmits the BFD echo packet to the peer system (1211). The transmission of the echo packet is terminated when there is no packet to be transmitted or when the session state is not UP ( 1212 ). Since the echo packet is transmitted through the BFD Echo Session (1213), it proceeds to step 1214. The BFD-Tx 1202 checks an adaptation completion flag to determine whether adaptation is finished (1214). When adaptation is finished, the BFD-Tx 1202 reads BFD Control session information and determined timer information (1209), creates a session based on the read session information and timer information (1210), and transmits a BFD Control packet (1211) ) BFD fault detection is performed. If the adaptation flag is disabled in step 1208, BFD session information and timer information are obtained without performing an adaptation operation (1209), a session is created (1210), and a BFD control packet is repeatedly transmitted (1211). In order to inform each other of the optimal timer determined through timer adaptation, the timer negotiation process is performed using the BFD poll sequence procedure. In addition to the operation of FIG. 5C , the BFD-Rx 1203 determines whether the received packet is a BFD control packet ( 1219 ). If it is not a BFD control packet, the BFD-Rx 1203 analyzes a loss result for the received echo packet (1224) and stores an adaptation timer (1225). When the adaptation is finished (1226), the BFD-Rx 1203 stores adaptation complete flags (1227). FIG. 13 shows a method for a device of the BFD system to transmit a control message. In step 1310, the first of the BFD system The device sequentially transmits a set of echo messages having different transmission timers to the second device. For example, it may correspond to sequentially transmitting a set of echo messages each having a different transmission period for a predetermined time interval. The first device transmits a first set of echo messages having a first transmission period, and when the first set of echo messages is looped back, it is determined whether transmission loss has occurred for the first set of echo messages, and a second A second set of echo messages having a transmission period may be transmitted. The first transmission period may correspond to, for example, the smallest period in which an echo message can be transmitted in a link between the first device and the second device. The second transmission period has a value smaller than the first transmission period if transmission loss for the first set of echo messages does not occur, and if transmission loss for the first set of echo messages occurs, the second transmission period It may have a value greater than one transmission period. The first device may determine the second transmission period in consideration of a current link utilization rate. Meanwhile, while the initial transmission loss of the echo message does not occur, the transmission period of the set of echo messages may increase exponentially.

In step 1320, the first device determines a transmission period of a control message based on a loopback result of the set of transmitted echo messages. For example, the first device may determine whether a preset condition is satisfied, and when the preset condition is satisfied, a transmission period of a set of echo messages transmitted when the preset condition is satisfied may be determined as the transmission period of the control message. If the preset condition is not satisfied, the first device may stop sequential transmission of the set of echo messages.

The preset condition is that a difference between a transmission period of a set of recent echo messages in which transmission loss occurs and a transmission period of a recent echo message in which transmission loss does not occur is less than a preset first reference value, or a set of echo messages in which transmission loss occurs. The number of transmissions is greater than the second preset reference value, or the number of transmissions of the set of echo messages in which no loss occurs is greater than the third reference value, or the transmission start time or transmission end time of the set of initially transmitted echo messages Corresponds to one of whether a first predetermined time has elapsed from, or a second predetermined time has elapsed from the transmission start time or transmission end time of the set of messages in which no loss has occurred for the first time, or a predetermined user input is detected. can do.

In step 1330, the first device transmits the control message to the second device according to the determined transmission period. The first device may transmit the control message according to the determined transmission period for a third predetermined time period. When the third predetermined time has elapsed, the first device sequentially transmits a set of echo messages to the second device, and restarts the transmission period of the control message based on a loopback result of the set of transmitted echo messages. and transmits the control message to the second device according to the determined transmission period.

14 shows the configuration of a BFD device.

Referring to FIG. 14 , the device of the BFD system may include the configuration of a communication unit 1410 , an input unit 1420 , a storage unit 1430 , and a control unit 1440 .

The communication unit 610 may transmit/receive data such as an echo message or a control message to and from the BFD device (system or peer). The input unit 1420 transmits a signal input by a user to the control unit 1440 in connection with function control of the device and setting of various functions. The input unit 1420 may be formed of a touch screen or a general keypad. The storage unit 1430 may record information collected by the device or information input by a user. The storage unit 1430 may store an application program corresponding to a function that can be performed in the device, data generated while a function is performed in the terminal, and the like. The storage unit 1430 may store, for example, information such as a transmission period of each device and an adaptation termination condition.

The controller 1440 controls the overall state and operation of units constituting the device. The controller 1440 may control other units to perform a function corresponding to the user input received through the input unit 1420 . The control unit 1440 may control to store the information received through the communication unit 1410 in the storage unit 1430 . The controller 1440 may control other units to perform various embodiments described in the present invention.

In FIG. 14 , the communication unit 1410 , the input unit 1420 , the storage unit 1430 , and the control unit 1440 are configured as separate blocks, and it has been described that each block performs different functions, but this is only for technical convenience. , each function is not necessarily separated in this way.

In the above-described embodiments, all steps may be selectively performed or omitted. Also, the steps in each embodiment do not necessarily occur in order and may be reversed.

The embodiments of the present invention disclosed in the present specification and drawings are merely provided for specific examples in order to easily explain the technical contents of the present invention and help the understanding of the present invention, and are not intended to limit the scope of the present invention.

Claims (20)

A method performed by a first node supporting bidirectional forwarding detection (BFD) in a communication system, the method comprising:
transmitting echo packets to the second node through a session between the first node and the second node based on a transmission period;
Based on a loop-back result of the echo packets, a first operation of increasing the transmission period to transmit the echo packets to the second node through the session or decreasing the transmission period to transmit the echo packets performing a second operation of transmitting to the second node through the session at least once until a preset termination condition is satisfied;
identifying the transmission period in which the first operation or the second operation is performed at least once until the preset termination condition is satisfied; and
and transmitting the echo packets to the second node through the session to detect failure of the session based on the identified transmission period.
According to claim 1,
The first operation is performed when it is identified that a loss has occurred based on the loopback result of the echo packets,
The second operation is performed when it is determined that the loss has not occurred based on the loopback result of the echo packets.
According to claim 1,
The transmission period of the initially transmitted echo packets is determined based on a packet transmission amount of the session.
According to claim 1,
The preset termination condition is, when a difference between a first transmission period of recent first echo packets in which loss has occurred and a second transmission period of recent second echo packets in which loss has not occurred is less than a first threshold, the When the number of the set of echo packets in which loss has occurred is equal to or greater than the second threshold, when the number of sets of echo packets in which no loss has occurred is equal to or greater than the third threshold, the transmission start time or the transmission end time of the initially transmitted echo packets and at least one of a case in which a first predetermined time has elapsed, or a case in which a second predetermined time has elapsed from a transmission start time or a transmission end time of the first lossless echo packets.
According to claim 1,
In the first operation, the width at which the transmission period is increased is determined based on the utilization rate of the session,
In the second operation, the width at which the transmission period is reduced is determined based on a utilization rate of the session.
According to claim 1,
In the first operation, the width at which the transmission period is increased is an average of the first transmission period of the most recent first echo packets in which the loss has occurred and the second transmission period of the recent second echo packets in which the loss does not occur. respond,
In the second operation, the width at which the transmission period is reduced corresponds to an average of the first transmission period and the second transmission period.
According to claim 1,
The method further comprising transmitting information on the identified transmission period to the second node through the session.
According to claim 1,
receiving, from the second node, the echo packets through the session;
The method of claim 1, further comprising: detecting occurrence of a failure of the session when the reception interval of the echo packets has a difference greater than or equal to a predetermined threshold value from the negotiated minimum reception interval.
In a first node supporting bidirectional forwarding detection (BFD) in a communication system, the first node comprises:
transceiver; and
A processor comprising:
Transmitting echo packets to the second node through a session between the first node and the second node based on a transmission period,
Based on a loop-back result of the echo packets, a first operation of increasing the transmission period to transmit the echo packets to the second node through the session or decreasing the transmission period to transmit the echo packets performing a second operation of transmitting to the second node through the session at least once until a preset termination condition is satisfied;
Identifies the transmission period in which the first operation or the second operation is performed at least once until the preset termination condition is satisfied;
and transmit the echo packets to the second node through the inter-session to detect occurrence of a failure for the session based on the identified transmission period.
10. The method of claim 9,
The first operation is performed when it is identified that a loss has occurred based on the loopback result of the echo packets,
The second operation is performed when it is determined that the loss has not occurred based on the loopback result of the echo packets.
10. The method of claim 9,
The first node, characterized in that the transmission period of the initially transmitted echo packets is determined based on a packet transmission amount of the session.
10. The method of claim 9,
The preset termination condition is, when a difference between a first transmission period of recent first echo packets in which loss has occurred and a second transmission period of recent second echo packets in which loss has not occurred is less than a first threshold, the When the number of the set of echo packets in which loss has occurred is equal to or greater than the second threshold, when the number of sets of echo packets in which no loss has occurred is equal to or greater than the third threshold, the transmission start time or the transmission end time of the initially transmitted echo packets The first node, characterized in that it includes at least one of a case in which a first predetermined time has elapsed, or a case in which a second predetermined time has elapsed from a transmission start time or a transmission end time of the first lossless echo packets. .
10. The method of claim 9,
In the first operation, the width at which the transmission period is increased is determined based on the utilization rate of the session,
In the second operation, a width at which the transmission period is reduced is determined based on a utilization rate of the session.
10. The method of claim 9,
In the first operation, the width at which the transmission period is increased is an average of the first transmission period of the most recent first echo packets in which the loss has occurred and the second transmission period of the recent second echo packets in which the loss does not occur. respond,
In the second operation, the width at which the transmission period is reduced corresponds to an average of the first transmission period and the second transmission period.
10. The method of claim 9, wherein the processor,
The first node, characterized in that the first node is further configured to transmit information on the identified transmission period to the second node through the session.
10. The method of claim 9, wherein the processor,
receive, from the second node, the echo packets through the session;
and detect occurrence of a failure of the session when the reception interval of the echo packets has a difference greater than or equal to a predetermined threshold value than the negotiated minimum reception interval. delete delete delete delete

Images