WO2013180287A1 - スイッチングシステム、ラインカード、スイッチカード、fdb学習方法、fdb学習調停方法及びプログラム - Google Patents
スイッチングシステム、ラインカード、スイッチカード、fdb学習方法、fdb学習調停方法及びプログラム Download PDFInfo
- Publication number
- WO2013180287A1 WO2013180287A1 PCT/JP2013/065245 JP2013065245W WO2013180287A1 WO 2013180287 A1 WO2013180287 A1 WO 2013180287A1 JP 2013065245 W JP2013065245 W JP 2013065245W WO 2013180287 A1 WO2013180287 A1 WO 2013180287A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- fdb
- learning
- fdb learning
- card
- learning information
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
- G06N5/027—Frames
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/54—Organization of routing tables
Definitions
- the present invention is based on a Japanese patent application: Japanese Patent Application No. 2012-126233 (filed on June 1, 2012), and the entire description of the application is incorporated herein by reference.
- the present invention relates to a switching system, a line card, a switch card, an FDB learning method, an FDB learning arbitration method, and a program, and in particular, a switching system, a line card, a switch card, and FDB learning that include a plurality of transfer databases (hereinafter, “FDB”).
- the present invention relates to a method, an FDB learning mediation method, and a program.
- Patent Document 1 includes a management unit that collects FDB information (FDB learning information) from a plurality of line cards (card A and card C of the same document) equipped with FDB and transmits them to each line card.
- FDB information FDB learning information
- a switch device comprising a second card provided is disclosed.
- Patent Document 2 discloses a configuration in which a transmission rate control circuit is mounted on each line card so that synchronization between FDBs is reliably performed.
- the transmission rate control circuit of Patent Document 2 controls the transmission rate of the learning frame by providing a time during which the learning frame is not transmitted forcibly for the time set in the counter (rate limiting section) (see FIG. 7, paragraph 0054). Assuming that no frame discard occurs, for example, when N line cards are connected, the transmission rate is (1 / (N ⁇ 1)) ⁇ 100%, and the time set in the counter is accordingly Will increase (see paragraph 0068). For this reason, there is a problem that only the worst number of FDBs can be learned even when there is a margin in the bandwidth of the relay route. According to the above formula, when 10 line cards are connected to a switch having 10 line card slots, the transmission rate is about 11.1%, and the FDB learning speed per unit time. Will be significantly slower.
- An object of the present invention is to provide a switching system, a line card, a switch card, an FDB learning method, an FDB learning arbitration method, and a program that can perform FDB learning at high speed without missing an FDB learning frame.
- a transfer database (hereinafter, “FDB”) and an entry that matches the received frame from the FDB are searched to determine whether or not FDB learning is necessary, and it is determined that FDB learning is necessary.
- the FDB learning unit that transmits FDB learning information to the switch card, and the FDB learning unit that registers the FDB learning information received from the switch card in the FDB.
- a plurality of line cards that notify whether or not they are in a possible state, a learning information buffer that accumulates FDB learning information received from the plurality of line cards, and all lines based on notifications from the line cards FDB learning information stored in the learning information buffer for each line card when the card is in a state where FDB learning is possible
- a switch card and a FDB learning arbitration unit selectively transmitting to the switching system
- a learning information buffer for accumulating FDB learning information received from a line card and a notification from the line card, when all line cards are in a state where FDB learning is possible, Whether or not FDB learning is required by searching for an entry that matches the received frame from the FDB, connected to a switch card including an FDB learning arbitration unit that transmits the accumulated FDB learning information to each line card. And when it is determined that FDB learning is necessary, an FDB search unit that transmits FDB learning information to the switch card and an FDB learning unit that registers the FDB learning information from the switch card in the FDB are provided.
- a line card for notifying the switch card whether or not FDB learning is possible is provided.
- a switch card based on a learning information buffer for accumulating FDB learning information received from a line card and a notification from the line card, when all line cards are in a state where FDB learning is possible, A switch card is provided that includes an FDB learning arbitration unit that transmits the accumulated FDB learning information to each line card.
- a line card connected to a switch card that includes the FDB learning arbitration unit that transmits the accumulated FDB learning information to each line card searches the FDB for an entry that matches the received frame, and requires FDB learning.
- a step of registering FDB learning information transmitted from a card in the FDB and an FDB learning method are provided.This method is linked to a specific machine called a line card that performs FDB learning.
- an entry that matches the received frame is searched from the FDB, it is determined whether or not FDB learning is necessary, and if it is determined that FDB learning is necessary, FDB learning is performed on the switch card.
- An FDB search unit that transmits information and an FDB learning unit that registers FDB learning information received from the switch card in the FDB, and notifies the switch card whether or not FDB learning is possible
- the switch card connected to the plurality of line cards stores the FDB learning information received from the line card and all the line cards are in a state capable of FDB learning based on the notification from the line card. Transmitting the accumulated FDB learning information to each of the line cards. It is.
- the method is tied to a specific machine called a switch card that sends FDB learning information to the line card.
- a computer program to be executed by a computer mounted on the above-described line card or switch card is provided.
- This program can be recorded on a computer-readable (non-transient) storage medium. That is, the present invention can be embodied as a computer program product.
- the present invention can be realized in a configuration in which a plurality of line cards 100 and 110 are connected to a switch card 200 as shown in FIG. More specifically, the line card 100 (110) searches the FDB 101 (111) and the FDB 101 (111) for an entry that matches the received frame, determines whether FDB learning is necessary, and learns the FDB. Is determined to be necessary, the FDB search unit 103 (113) that transmits FDB learning information to the switch card 200, and the FDB learning unit 102 (registers) the FDB learning information received from the switch card 200 in the FDB 101 (111). 112). Further, each line card 100, 110 performs an operation of notifying the switch card 200 whether or not the own device is in a state where FDB learning is possible.
- the switch card 200 is in a state in which all line cards can perform FDB learning based on a learning information buffer 201 that accumulates FDB learning information received from the line cards 100 and 110, and notifications from the line cards 100 and 110.
- an FDB learning arbitration unit 202 that selects and transmits FDB learning information stored in the learning information buffer 201 to each of the line cards 100 and 110 is provided.
- FIG. 1 is a diagram illustrating a configuration of a switching system according to a first embodiment of this invention.
- a configuration in which a plurality of line cards 100 and 110 are connected to a switch card 200 is shown.
- the frame transfer process using the FDB 101 (111) is omitted.
- the line card 100 (110) includes an FDB 101 (111), an FDB search unit 103 (113), and an FDB learning unit 102 (112).
- the FDB 101 (111) is a database that stores information necessary for transferring Ethernet frames.
- the FDB 101 (111) can also be realized by a table or the like that stores 0 sets or multiple sets of MAC addresses and line numbers as one set.
- the line card 100 (110) is provided with a plurality of ports, each entry of the FDB 101 (111) is associated with a port number.
- the FDB 101 (111) When the FDB 101 (111) receives an inquiry from the FDB search unit 103 (113), the FDB 101 (111) searches the corresponding entry using the MAC address or the like as a key, and transmits the search result to the FDB search unit 103 (113). Further, the FDB 101 (111) receives registration of FDB learning information from the FDB learning unit 102 (112).
- the FDB learning unit 102 (112) is connected to the FDB 101 (111) and the FDB learning arbitration unit 202 of the switch card 200.
- the FDB learning unit 102 (112) receives FDB learning information from the FDB learning arbitration unit 202 of the switch card 200, the FDB learning unit 102 (112) registers the FDB learning information in the FDB 101 (111).
- the FDB learning unit 102 (112) sends back pressure (for example, to the FDB 101 (111)) as information indicating whether the FDB learning arbitration unit 202 of the switch card 200 is in a state where FDB learning is possible.
- the first value (High) is transmitted when the access interface is in use, and the second value (low) is transmitted when the access interface is not used.
- information indicating whether or not the FDB learning is possible information on the current load state or the like may be transmitted to the FDB learning arbitration unit 202 of the switch card 200 in addition to back pressure. For example, when the CPU usage rate exceeds 90%, a packet indicating the occurrence of the CPU congestion state is transmitted, and when the CPU usage rate falls below 90%, a packet indicating the release of the CPU congestion state can be transmitted.
- the FDB search unit 103 (113) is connected to the outside, the FDB 101 (111), and the FDB learning arbitration unit 202 of the switch card 200.
- FDB search unit 103 (113) receives a frame from the outside, FDB learning information (source MAC address and line number of the received frame) and search information (MAC address of the received frame) are received from the received frame and the received line. Create and query the FDB 101 (111) for an entry corresponding to the search information.
- the FDB search unit 103 (113) determines that FDB learning is necessary, FDB learning information is transmitted to the FDB learning arbitration unit 202 of the switch card 200.
- the FDB search unit 103 (113) does not need to learn the FDB. Judge and do nothing.
- the switch card includes a learning information buffer 201 and an FDB learning arbitration unit 202.
- the FDB learning arbitration unit 202 is connected to the learning information buffer 201, the FDB search unit 103 (113) of the line card 100 (110), and the FDB learning unit 102 (112). Upon receiving FDB learning information from the FDB search unit 103 (113), the FDB learning arbitration unit 202 stores it in the learning information buffer 201. Further, when the back pressure is not received from all the FDB learning units 102 and 112, the FDB learning arbitration unit 202 transmits the FDB learning information extracted from the learning information buffer 201 to all the line cards 100 and 110. To do. Note that the selection criteria for FDB learning information to be extracted when a plurality of FDB learning information is stored in the learning information buffer 201 can be arbitrarily set. On the other hand, when back pressure is received from any of the FDB learning units 102 and 112, the FDB learning arbitration unit 202 pauses transmission of FDB learning information.
- the FDB learning arbitration unit 202 whether or not the back pressure is received from all the FDB learning units 102 and 112 is determined based on signals transmitted from the line cards 100 and 110 (for example, a learning state if High). , It is possible to realize a configuration in which the state of all the line cards is grasped using a logical sum or logical product of a state in which learning is impossible.
- the learning information buffer 201 stores the FDB learning information transmitted from the FDB learning arbitration unit 202.
- the FDB learning arbitration unit 202 deletes the FDB learning information read from the learning information buffer 201.
- the units (processing means) of the line card and the switch card shown in FIG. 1 are realized by a computer program that causes a computer mounted on these devices to execute the above-described processes using the hardware. You can also.
- FIG. 3 is a flowchart showing the flow of FDB search processing performed when a frame is received by the line card 100 (110) according to the first embodiment of this invention.
- the line card 100 (110) receives a frame from the outside by the FDB search unit 103 (113)
- the line card 100 (110) creates FDB learning information and search information from the received frame and the reception line.
- the FDB search unit 103 (113) inquires of the FDB 101 (111) whether there is a corresponding entry using the search information as a key.
- the FDB 101 (111) searches the table and the like illustrated in FIG. 2, and returns a search result to the FDB search unit 103 (113) (step S201).
- the FDB search unit 103 stores in the FDB learning arbitration unit 202 of the switch card 200. On the other hand, FDB learning information is transmitted (step S204), and the process ends.
- the FDB search unit 103 receives the entry returned from the FDB 101 (111) and FDB learning information (The transmission source MAC address and line number of the received frame are compared (step S203).
- step S203 if the two match (Yes in step S203), the FDB search unit 103 (113) does nothing and ends the process.
- the FDB search unit 103 transmits FDB learning information to the FDB learning arbitration unit 202 of the switch card 200 (step S204). The process is terminated.
- FIG. 4 is a flowchart showing a flow of storage processing performed when FDB learning information is received by the switch card 200 according to the first embodiment of this invention.
- the FDB learning arbitration unit 202 of the switch card 200 upon receiving FDB learning information from the FDB search unit 103 (113) of the line card 100 (110), stores the FDB learning information in the learning information buffer 201. (Step S301), the process ends. Note that the storage position of the FDB learning information in the learning information buffer 201 is determined according to a predetermined rule. In the present embodiment, new FDB learning information is stored at the end of the learning information buffer 201.
- FIG. 5 is a flowchart showing the flow of FDB learning information transmission processing (arbitration processing) of the switch card 200 of this embodiment.
- the “arbitration process” refers to a process of transmitting FDB learning information to the FDB learning unit 102 (112) at a timing at which FDB learning can be reliably performed.
- step S401 when the FDB learning arbitration unit 202 of the switch card 200 confirms that one or more FDB learning information is stored in the learning information buffer 201, the FDB learning unit 102 (112) of each line card Then, it is determined whether or not the back pressure is received (step S401).
- the FDB learning arbitration unit 202 postpones subsequent transmission processing of the FDB learning information (returns to step S401).
- the FDB learning arbitration unit 202 extracts the FDB learning information in the front row from the learning information buffer 201 (step S402).
- the FDB learning arbitration unit 202 transmits the extracted FDB learning information to the FDB learning units 102 (112) of all the connected line cards 100 (110) (step S403), and ends the processing. To do. Note that after the step S402 or S403 is completed, the transmitted FDB learning information is deleted from the learning information buffer 201.
- FIG. 6 is a flowchart showing the operation of the line card 100 (110) of this embodiment when receiving FDB learning information.
- the FDB learning unit 102 (112) of the line card 100 (110) receives FDB learning information from the FDB learning arbitration unit 202 of the switch card 200
- the FDB learning arbitration unit 202 of the switch card 200 receives the FDB learning arbitration unit 202.
- transmission of back pressure is started (step S501). The back pressure is continuously transmitted until registration in the FDB 101 (111) is completed.
- the FDB learning unit 102 (112) registers the FDB learning information in the FDB 101 (111) in the same line card (self apparatus) (step S502).
- the FDB learning unit 102 (112) stops transmitting back pressure to the FDB learning arbitration unit 202 (step S503).
- the FDB learning unit 102 based on the back pressure transmitted by the FDB learning unit 102 (112), it is determined whether or not all line cards are in a state where FDB learning is possible. For this reason, when all the line cards are in a state where FDB learning is possible, the learning process can be actively performed. For this reason, resources can be fully utilized, and high-speed FDB learning is realized.
- FIG. 7 is a diagram illustrating a configuration of a switching system according to the second embodiment of this invention.
- the line card 120 (130) side processes a frame for transmitting and receiving a frame (see FIG. 8; hereinafter referred to as “dedicated frame”) with the switch card 210.
- a processing unit 124 (134) is provided, a flag processing unit 213 that processes a flag included in a dedicated frame sent from the frame processing unit 124 (134) on the switch card 210 side, and a line card 120 ( 130), a frame transfer processing unit 214 for transferring a dedicated frame is provided.
- a frame transfer processing unit 214 for transferring a dedicated frame is provided.
- the FDB search unit 123 (133) of the line card 120 (130) of the present embodiment transmits the frame received from the outside, the FDB learning information, and the flag to the frame processing unit.
- the flag a value indicating whether or not FDB learning is necessary is set, and “1” is set when it is determined that FDB learning is necessary, and “0” is set otherwise.
- the FDB search unit 123 (133) of the line card 120 (130) of the present embodiment does not transmit the frame to the outside, but only receives it.
- the frame processing unit 124 (134) Upon receiving the FDB learning information, the frame, and the flag from the FDB search unit 123 (133), the frame processing unit 124 (134) adds an additional header that stores the FDB learning information (line number) and the flag to the Ethernet frame.
- a dedicated frame (see FIG. 8) is created, and the created dedicated frame is transmitted to the flag processing unit 213 of the switch card 210. Further, when the frame processing unit 124 (134) receives the dedicated frame from the frame transfer processing unit 214 of the switch card 210, the frame processing unit 124 (134) deletes the FDB learning information and the flag from the dedicated frame and transmits them to the outside.
- the FDB 121 (131) and the FDB learning unit 122 (132) of the line card 120 (130) are equivalent to the FDB 101 (111) and the FDB learning unit 102 (112) of the first embodiment, the description thereof is omitted. To do.
- the FDB learning arbitration unit 212 of the switch card 210 receives FDB learning information from the flag processing unit 213 instead of receiving FDB learning information directly from the line card 120 (130). In addition, the operation when the FDB learning information is received is the same.
- the flag processing unit 213 When the flag processing unit 213 receives the dedicated frame from the frame processing unit 124 (134) of the line card 120 (130), the flag processing unit 213 confirms the flag in the dedicated frame and determines subsequent processing. When the flag is 1 (learning required), the flag processing unit 213 extracts FDB learning information (source MA address and line number) from the dedicated frame, and transmits the FDB learning information to the FDB learning arbitration unit 212. At the same time, a dedicated frame is transmitted to the frame transfer processing unit 214. On the other hand, when the flag is 0 (no learning is necessary), the flag processing unit 213 performs processing for transmitting a dedicated frame to the frame transfer processing unit 214.
- FDB learning information source MA address and line number
- the frame transfer processing unit 214 When the frame transfer processing unit 214 receives the dedicated frame from the flag processing unit 213, the frame transfer processing unit 214 checks the flag in the dedicated frame and determines subsequent processing. When the flag is 1 (learning required), the frame transfer processing unit 214 performs processing for transmitting a dedicated frame to the frame processing unit of the line card corresponding to the line card number in the dedicated frame. On the other hand, when the flag is 0 (no learning is required), the frame transfer processing unit 214 transmits a dedicated frame to the frame processing units 124 (134) of all the line cards.
- the learning information buffer 211 is the same as the learning information buffer 201 of the first embodiment, and a description thereof will be omitted.
- FIG. 9 is a flowchart showing the flow of FDB search processing performed when a frame is received by the line card 120 (130) according to the second embodiment of this invention.
- the line card 120 (130) receives a frame from the outside at the FDB search unit 123 (133)
- the line card 120 (130) creates FDB learning information and search information from the received frame and the reception line.
- the FDB search unit 123 (133) inquires of the FDB 121 (131) whether there is a corresponding entry using the search information as a key.
- the FDB 121 (131) searches the table and the like illustrated in FIG. 2, and returns a search result to the FDB search unit 123 (133) (step S701).
- the FDB search unit 123 if an entry corresponding to the corresponding MAC address is not registered in the FDB 121 (131) (No in step S702), the FDB search unit 123 (133) sends a request to the frame processing unit 124 (134).
- step S703 when an entry corresponding to the corresponding MAC address is registered in the FDB 121 (131) (Yes in step S702), the FDB search unit 123 (133) performs the FDB learning with the entry returned from the FDB 121 (131).
- the information (the source MAC address of the received frame and the line number) is compared (step S703).
- the frame processing unit 124 uses the flag received from the FDB search unit 123 (133), the FDB learning information, and the received frame to create a dedicated frame with the additional header shown in FIG. , To the flag processing unit 213 of the switch card 210.
- the necessity of FDB learning is determined every time a frame is input to the line card. Then, an operation of transmitting a dedicated frame storing a flag indicating the result and line information to the switch card 210 is performed.
- FIG. 10 is a flowchart showing the flow of processing performed when a dedicated frame is received in the switch card 210 according to the second embodiment of this invention.
- the flag processing unit 213 of the switch card 210 when the flag processing unit 213 of the switch card 210 receives the dedicated frame, it checks the flag in the dedicated frame (step S801).
- the flag processing unit 213 extracts the FDB learning information (source MAC address, line number) from the dedicated frame, and the FDB learning arbitration unit 212 receives the FDB Learning information is transmitted (step S803).
- the FDB learning arbitration unit 212 When receiving the FDB learning information, stores the received FDB learning information in the learning information buffer 211 (step S804).
- the flag processing unit 213 of the switch card 210 transmits a dedicated frame to the frame transfer processing unit 214 and causes all the line cards to transfer the dedicated frame (step S805).
- the flag processing unit 213 refers to the line number in the dedicated frame and transmits the dedicated frame to the corresponding line card. (Step S806).
- (3A) FDB learning arbitration processing (4A) FDB learning processing
- (3A) FDB learning arbitration processing and (4A) FDB learning processing are the same as those in the first embodiment, and a description thereof will be omitted ( (See FIGS. 5 and 6).
- FIG. 11 is a flowchart showing the flow of processing performed when a dedicated frame is received by the line card 120 (130) according to the second embodiment of this invention.
- step S1101 When the dedicated frame transmitted in steps S805 and S806 in FIG. 10 is input to the frame processing unit 124 (134) of the line card 120 (130), the frame processing unit 124 (134), as shown in FIG.
- the additional header storing the flag and line number information is deleted from the received dedicated frame, returned to the normal Ethernet frame, and the frame is transferred to the outside (step S1101).
- This embodiment has the following effects in addition to the same effects as those of the first embodiment.
- the line card and the switch card are generally connected by a connector so that the line card can be exchanged. Since this connector has a finite number of pins, in order to wire a signal exceeding the allowable number of pins of the connector between the line card and the switch card, it is necessary to increase the number of connectors.
- the increase in the number of connectors may increase the mounting area of the board and increase the cost of connector parts. Therefore, the designer spends a lot of time on designing to exchange a lot of information with a small number of connectors (number of signals).
- the frame and the learning information are processed by different routes, and a physical signal line is provided separately from the frame communication route for the FDB learning information.
- a physical signal line is provided separately from the frame communication route for the FDB learning information.
- the line card is described using the back pressure to notify the switch card whether or not FDB learning is possible, but other signals are used. It is also possible to notify whether or not FDB learning is possible. Of course, instead of the FDB learning unit, another processing means may notify whether or not FDB learning is possible.
- the line card further includes a frame processing unit that stores information to be transmitted to the switch card in a predetermined frame, The switching system, wherein the switch card extracts FDB learning information from a frame based on a flag included in the frame and registers the information in the learning information storage unit.
- FDB learning information In the switching system of the fourth form, A switching system in which the received frame and the FDB learning information are transmitted through the same physical wiring using the predetermined frame.
- the switch card starts transmission of FDB learning information to the switch card based on a logical sum or logical product of signals indicating whether or not the FDB learning is received from each line card.
- FDB transfer database
- An FDB search unit that transmits information and an FDB learning unit that registers FDB learning information received from the switch card in the FDB, and notifies the switch card whether or not FDB learning is possible
- a computer mounted on a switch card connected to multiple line cards A process of accumulating FDB learning information received from the line card; Based on the notification from the line card, when all line cards are in a state where FDB learning is possible, a process of transmitting the accumulated FDB learning information to each line card;
- a program that executes
Abstract
Description
本発明は、日本国特許出願:特願2012-126233号(2012年6月1日出願)に基づくものであり、同出願の全記載内容は引用をもって本書に組み込み記載されているものとする。
本発明は、スイッチングシステム、ラインカード、スイッチカード、FDB学習方法、FDB学習調停方法及びプログラムに関し、特に、転送データベース(以下、「FDB」)を複数備えるスイッチングシステム、ラインカード、スイッチカード、FDB学習方法、FDB学習調停方法及びプログラムに関する。
続いて、本発明の第1の実施形態について図面を参照して詳細に説明する。図1は、本発明の第1の実施形態のスイッチングシステムの構成を示す図である。図1を参照すると、スイッチカード200に、複数のラインカード100、110を接続した構成が示されている。以下の説明では、1つのスイッチカードに4枚のラインカードが接続されているものとし、1つのラインカードに1つのイーサネット(登録商標)ポートがあるものとする。また、本実施形態では、FDB101(111)を用いたフレームの転送処理については省略する。
図3は、本発明の第1の実施形態のラインカード100(110)のフレーム受信時に行われるFDB検索処理の流れを表したフローチャートである。
図4は、本発明の第1の実施形態のスイッチカード200のFDB学習情報の受信時に行われる格納処理の流れを表したフローチャートである。
図5は、本実施形態のスイッチカード200のFDB学習情報送信処理(調停処理)の流れを表したフローチャートである。ここで、「調停処理」は、確実にFDB学習を実施可能なタイミングで、FDB学習部102(112)に対して、FDB学習情報を送信する処理を指している。
図6は、本実施形態のラインカード100(110)のFDB学習情報受信時の動作を表したフローチャートである。図6を参照すると、ラインカード100(110)のFDB学習部102(112)は、スイッチカード200のFDB学習調停部202からFDB学習情報を受信すると、スイッチカード200のFDB学習調停部202に対して、バックプレッシャーの送信を開始する(ステップS501)。なお、バックプレッシャーは、FDB101(111)への登録が完了するまで、継続して送信される。
続いて、ラインカードとスイッチカード間のFDB学習情報等の授受をフレームと同一ルートで行うようにした本発明の第2の実施形態について説明する。図7は、本発明の第2の実施形態のスイッチングシステムの構成を示す図である。
図9は、本発明の第2の実施形態のラインカード120(130)のフレーム受信時に行われるFDB検索処理の流れを表したフローチャートである。
図10は、本発明の第2の実施形態のスイッチカード210において専用フレームの受信時に行われる処理の流れを表したフローチャートである。
(4A)FDB学習処理
本実施形態における(3A)FDB学習の調停処理及び(4A)FDB学習処理は、第1の実施形態と同様であるので説明を省略する(図5、図6参照)。
図11は、本発明の第2の実施形態のラインカード120(130)の専用フレーム受信時に行われる処理の流れを表したフローチャートである。
[第1の形態]
(上記第1の視点によるスイッチングシステム参照)
[第2の形態]
第1の形態のスイッチングシステムにおいて、
前記ラインカードは、前記FDBに前記FDB学習情報を書き込んでいる期間は、前記スイッチカードに対してFDB学習が可能な状態でないことを通知するスイッチングシステム。
[第3の形態]
第1又は第2の形態のスイッチングシステムにおいて、
前記ラインカードは、自装置の負荷状態を表す数値が所定のしきい値を超えたか否かにより、FDB学習が可能な状態にあるかを通知するスイッチングシステム。
[第4の形態]
第1から第3いずれか一の形態のスイッチングシステムにおいて、
前記ラインカードは、さらに、前記スイッチカードに対して送信する情報を所定のフレームに格納するフレーム処理部を備え、
前記スイッチカードは、前記フレームに含まれるフラグに基づいてフレームからFDB学習情報を取り出して前記学習情報蓄積部に登録するスイッチングシステム。
[第5の形態]
第4の形態のスイッチングシステムにおいて、
前記所定のフレームを用いて、前記受信フレームと前記FDB学習情報とが同一の物理的配線で送信されるスイッチングシステム。
[第6の形態]
第1から第5いずれか一の形態のスイッチングシステムにおいて、
前記スイッチカードは、前記各ラインカードから受信したFDB学習が可能な状態にあるか否かを示す信号の論理和又は論理積に基づいて、前記スイッチカードに対するFDB学習情報を送信を開始するスイッチングシステム。
[第7の形態]
(上記第2の視点によるラインカード参照)
[第8の形態]
(上記第3の視点によるスイッチカード参照)
[第9の形態]
(上記第4の視点によるFDB学習方法参照)
[第10の形態]
(上記第5の視点によるFDB学習調停方法参照)
[第11の形態]
ラインカードから受信した転送データベース(以下、「FDB」)学習情報を蓄積する学習情報蓄積部と、前記ラインカードからの通知に基づいて、すべてのラインカードがFDB学習可能な状態であるときに、前記各ラインカードに対して前記蓄積したFDB学習情報を送信するFDB学習調停部とを備えるスイッチカードと接続されたラインカードに搭載されたコンピュータに、
FDBから受信フレームに適合するエントリを検索して、FDBの学習が必要か否かを判定する処理と、
FDBの学習が必要と判定した場合、スイッチカードに対してFDB学習情報を送信する処理と、
すべてのラインカードがFDB学習可能な状態であるときに、前記スイッチカードから送信されるFDB学習情報を前記FDBに登録する処理と、を実行させるプログラム。
[第12の形態]
転送データベース(以下、「FDB」)から受信フレームに適合するエントリを検索して、FDBの学習が必要か否かを判定し、FDBの学習が必要と判定した場合、スイッチカードに対してFDB学習情報を送信するFDB検索部と、前記スイッチカードから受信したFDB学習情報を前記FDBに登録するFDB学習部とを備え、前記スイッチカードに対し、FDB学習が可能な状態にあるか否かを通知する複数のラインカードと接続されたスイッチカードに搭載されたコンピュータに、
ラインカードから受信したFDB学習情報を蓄積する処理と、
前記ラインカードからの通知に基づいて、すべてのラインカードがFDB学習可能な状態であるときに、前記各ラインカードに対して前記蓄積したFDB学習情報を送信する処理と、
を実行させるプログラム。
101、111、121、131 FDB
102、112、122、132 FDB学習部
103、113、123、133 FDB検索部
124、134 フレーム処理部
200、210 スイッチカード
201、211 学習情報バッファ
202、212 FDB学習調停部
213 フラグ処理部
214 フレーム転送処理部
Claims (10)
- 転送データベース(以下、「FDB」)と、FDBから受信フレームに適合するエントリを検索してFDBの学習が必要か否かを判定し、FDBの学習が必要と判定した場合、スイッチカードに対してFDB学習情報を送信するFDB検索部と、前記スイッチカードから受信したFDB学習情報を前記FDBに登録するFDB学習部とを備え、前記スイッチカードに対し、FDB学習が可能な状態にあるか否かを通知する複数のラインカードと、
前記複数のラインカードから受信したFDB学習情報を蓄積する学習情報蓄積部と、前記各ラインカードからの通知に基づいて、すべてのラインカードがFDB学習可能な状態であるときに、前記各ラインカードに対して前記学習情報蓄積部に蓄積したFDB学習情報を選択して送信するFDB学習調停部とを備えるスイッチカードと、を含むこと、
を特徴とするスイッチングシステム。 - 前記ラインカードは、前記FDBに前記FDB学習情報を書き込んでいる期間は、前記スイッチカードに対してFDB学習が可能な状態でないことを通知する請求項1のスイッチングシステム。
- 前記ラインカードは、自装置の負荷状態を表す数値が所定のしきい値を超えたか否かにより、FDB学習が可能な状態にあるかを通知する請求項1又は2のスイッチングシステム。
- 前記ラインカードは、さらに、前記スイッチカードに対して送信する情報を所定のフレームに格納するフレーム処理部を備え、
前記スイッチカードは、前記フレームに含まれるフラグに基づいてフレームからFDB学習情報を取り出して前記学習情報蓄積部に登録する請求項1から3いずれか一のスイッチングシステム。 - 前記所定のフレームを用いて、前記受信フレームと前記FDB学習情報とが同一の物理的配線で送信される請求項4のスイッチングシステム。
- 前記スイッチカードは、前記各ラインカードから受信したFDB学習が可能な状態にあるか否かを示す信号の論理和又は論理積に基づいて、前記スイッチカードに対するFDB学習情報を送信を開始する請求項1から5いずれか一のスイッチングシステム。
- ラインカードから受信した転送データベース(以下、「FDB」)学習情報を蓄積する学習情報蓄積部と、前記ラインカードからの通知に基づいて、すべてのラインカードがFDB学習可能な状態であるときに、前記各ラインカードに対して前記蓄積したFDB学習情報を送信するFDB学習調停部とを備えるスイッチカードと、接続され、
前記FDBから受信フレームに適合するエントリを検索して、FDBの学習が必要か否かを判定し、FDBの学習が必要と判定した場合、スイッチカードに対してFDB学習情報を送信するFDB検索部と、前記スイッチカードから受信したFDB学習情報を前記FDBに登録するFDB学習部とを備え、前記スイッチカードに対し、FDB学習が可能な状態にあるか否かを通知するラインカード。 - ラインカードから受信した転送データベース(以下、「FDB」)学習情報を蓄積する学習情報蓄積部と、前記ラインカードからの通知に基づいて、すべてのラインカードがFDB学習可能な状態であるときに、前記各ラインカードに対して前記蓄積したFDB学習情報を送信するFDB学習調停部とを備えるスイッチカード。
- ラインカードから受信した転送データベース(以下、「FDB」)学習情報を蓄積する学習情報蓄積部と、前記ラインカードからの通知に基づいて、すべてのラインカードがFDB学習可能な状態であるときに、前記各ラインカードに対して前記蓄積したFDB学習情報を送信するFDB学習調停部とを備えるスイッチカードと接続されたラインカードが、
FDBから受信フレームに適合するエントリを検索して、FDBの学習が必要か否かを判定するステップと、
FDBの学習が必要と判定した場合、スイッチカードに対してFDB学習情報を送信するステップと、
すべてのラインカードがFDB学習可能な状態であるときに、前記スイッチカードから送信されるFDB学習情報を前記FDBに登録するステップと、を含むFDB学習方法。 - 転送データベース(以下、「FDB」)から受信フレームに適合するエントリを検索して、FDBの学習が必要か否かを判定し、FDBの学習が必要と判定した場合、スイッチカードに対してFDB学習情報を送信するFDB検索部と、前記スイッチカードから受信したFDB学習情報を前記FDBに登録するFDB学習部とを備え、前記スイッチカードに対し、FDB学習が可能な状態にあるか否かを通知する複数のラインカードと接続されたスイッチカードが、
ラインカードから受信したFDB学習情報を蓄積するステップと、
前記ラインカードからの通知に基づいて、すべてのラインカードがFDB学習可能な状態であるときに、前記各ラインカードに対して前記蓄積したFDB学習情報を送信するステップと、
を含むFDB学習調停方法。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201380028892.XA CN104350709B (zh) | 2012-06-01 | 2013-05-31 | 交换机***、线路卡、交换卡、fdb学习方法及fdb学习调停方法 |
US14/404,642 US9792559B2 (en) | 2012-06-01 | 2013-05-31 | Switching system, line card, switch card, FDB learning method, FDB learning arbitration method and program |
JP2014518762A JP6079775B2 (ja) | 2012-06-01 | 2013-05-31 | スイッチングシステム、ラインカード、スイッチカード、fdb学習方法、fdb学習調停方法及びプログラム |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2012-126233 | 2012-06-01 | ||
JP2012126233 | 2012-06-01 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2013180287A1 true WO2013180287A1 (ja) | 2013-12-05 |
Family
ID=49673470
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2013/065245 WO2013180287A1 (ja) | 2012-06-01 | 2013-05-31 | スイッチングシステム、ラインカード、スイッチカード、fdb学習方法、fdb学習調停方法及びプログラム |
Country Status (4)
Country | Link |
---|---|
US (1) | US9792559B2 (ja) |
JP (1) | JP6079775B2 (ja) |
CN (1) | CN104350709B (ja) |
WO (1) | WO2013180287A1 (ja) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
IL297591A (en) | 2014-04-10 | 2022-12-01 | Seattle Childrens Hospital Dba Seattle Childrens Res Inst | Drug-related transgene expression |
CN108174604B (zh) | 2015-08-07 | 2023-06-23 | 西雅图儿童医院(Dba西雅图儿童研究所) | 用于实体瘤靶向的双特异性car t细胞 |
CN107872340B (zh) * | 2016-09-27 | 2021-09-07 | 华为技术有限公司 | 一种单板注册的方法、单板及转发设备 |
AU2017375630B2 (en) | 2016-12-12 | 2023-12-21 | Seattle Children's Hospital (dba Seattle Children's Research Institute) | Chimeric transcription factor variants with augmented sensitivity to drug ligand induction of transgene expression in mammalian cells |
CN110661733B (zh) * | 2018-06-29 | 2021-07-30 | 智邦科技股份有限公司 | 网络接口卡、交换卡及线路卡 |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006237801A (ja) * | 2005-02-23 | 2006-09-07 | Fujitsu Access Ltd | Adslサービスシステム及びadslサービス速度切替方法 |
JP2010177722A (ja) * | 2009-01-27 | 2010-08-12 | Nec Corp | スイッチ装置、カード、fdb情報の管理方法及びプログラム |
JP2011091477A (ja) * | 2009-10-20 | 2011-05-06 | Hitachi Cable Ltd | スイッチングハブ及びfdbの同期方法 |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100592873B1 (ko) * | 2002-12-20 | 2006-06-23 | 한국전자통신연구원 | 복수개의 프로세서 및 다중/역다중화기를 갖는 라인카드에서 주소학습에 따른 네트워크 프로세서의 포워딩 테이블 관리방법 |
US7593400B2 (en) * | 2006-05-19 | 2009-09-22 | Corrigent Systems Ltd. | MAC address learning in a distributed bridge |
US7660303B2 (en) * | 2006-08-22 | 2010-02-09 | Corrigent Systems Ltd. | Point-to-multipoint functionality in a bridged network |
CN101488862B (zh) * | 2009-02-23 | 2012-02-08 | 中兴通讯股份有限公司 | 分布式以太网交换机及其内部的mac地址维护方法 |
CN101547147B (zh) * | 2009-04-30 | 2012-01-25 | 北京星网锐捷网络技术有限公司 | 一种报文处理方法、装置、线卡和管理板 |
CN102045200A (zh) * | 2010-12-24 | 2011-05-04 | 武汉烽火网络有限责任公司 | 分布式交换机mac地址表项有条件同步的方法 |
-
2013
- 2013-05-31 US US14/404,642 patent/US9792559B2/en not_active Expired - Fee Related
- 2013-05-31 JP JP2014518762A patent/JP6079775B2/ja not_active Expired - Fee Related
- 2013-05-31 WO PCT/JP2013/065245 patent/WO2013180287A1/ja active Application Filing
- 2013-05-31 CN CN201380028892.XA patent/CN104350709B/zh not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006237801A (ja) * | 2005-02-23 | 2006-09-07 | Fujitsu Access Ltd | Adslサービスシステム及びadslサービス速度切替方法 |
JP2010177722A (ja) * | 2009-01-27 | 2010-08-12 | Nec Corp | スイッチ装置、カード、fdb情報の管理方法及びプログラム |
JP2011091477A (ja) * | 2009-10-20 | 2011-05-06 | Hitachi Cable Ltd | スイッチングハブ及びfdbの同期方法 |
Also Published As
Publication number | Publication date |
---|---|
US9792559B2 (en) | 2017-10-17 |
JP6079775B2 (ja) | 2017-02-15 |
CN104350709B (zh) | 2018-03-13 |
JPWO2013180287A1 (ja) | 2016-01-21 |
US20150120622A1 (en) | 2015-04-30 |
CN104350709A (zh) | 2015-02-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6079775B2 (ja) | スイッチングシステム、ラインカード、スイッチカード、fdb学習方法、fdb学習調停方法及びプログラム | |
US8358658B2 (en) | Implementing ordered and reliable transfer of packets while spraying packets over multiple links | |
CN101150527B (zh) | 一种pcie数据传输的方法、***及设备 | |
US8543754B2 (en) | Low latency precedence ordering in a PCI express multiple root I/O virtualization environment | |
US11228488B2 (en) | Software implementation of network switch/router | |
JP5104465B2 (ja) | 転送装置及びパケット伝送装置 | |
US9455916B2 (en) | Method and system for changing path and controller thereof | |
JP2008079056A (ja) | 通信装置及びそれに用いる検索テーブル管理方法 | |
US8340112B2 (en) | Implementing enhanced link bandwidth in a headless interconnect chip | |
JP5787061B2 (ja) | スイッチシステム、ラインカード、fdb情報の学習方法及びプログラム | |
JP4667849B2 (ja) | ループ検出方法及び装置 | |
JP2009246524A (ja) | 信号伝送装置 | |
US7564860B2 (en) | Apparatus and method for workflow-based routing in a distributed architecture router | |
US20160205013A1 (en) | Transmission and Reception Devices | |
US20140304450A1 (en) | Switching device, packet control method, and data communication system | |
US10728178B2 (en) | Apparatus and method for distribution of congestion information in a switch | |
US10609188B2 (en) | Information processing apparatus, information processing system and method of controlling information processing system | |
JP2015170948A (ja) | 情報処理システム、情報処理装置、スイッチ装置および情報処理システムの制御方法 | |
US5787081A (en) | Allocation of node transmissions in switching networks used with scalable coherent interfaces | |
US20150180775A1 (en) | Communication System, Control Apparatus, Communication Method, and Program | |
CN108123990B (zh) | 一种数据存储方法、数据存储***及数据处理设备 | |
JP4655733B2 (ja) | リング型二重化ネットワークにおける受信方法及び装置 | |
US11038800B2 (en) | Techniques for reducing the overhead of providing responses in a computing network | |
US10855609B2 (en) | Interconnect and method of operation of such an interconnect | |
US20150317251A1 (en) | Maintaining a system state cache |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 13798054 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2014518762 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14404642 Country of ref document: US |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 13798054 Country of ref document: EP Kind code of ref document: A1 |