JP2013005143A - Ring type network system, network management apparatus, and layer 2 switch - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a ring type network system having high accommodation efficiency.SOLUTION: In a ring type network including a plurality of layer 2 switches 11-14 with a ring protection mounted thereon, a first path and a second path between a first layer 2 switch and a second layer 2 switch are calculated so as to avoid mutual overlapping physical links, thereby assigning a first virtual network identifier #A corresponding to the first path, and a second virtual network identifier #B corresponding to the second path. Each of the layer 2 switches 11-14 extracts a virtual network identifier from a frame, and performs switching process of the frame on the basis of setting information corresponding to the virtual network identifier.

Description

  The present invention relates to a ring network system having redundancy by connecting L2 switches in a ring shape, and more particularly to a frame communication path on the network.

  The Ethernet (registered trademark) technology is standardized by IEEE 802.3 as a definition of a physical layer / data link layer in a LAN (Local Area Network), and is widely used in a local area LAN or the like. In recent years, it has been proposed to apply the Ethernet technology to a wide area network (WAN) such as a backbone network or a relay network in a carrier. In such applications, high fault tolerance is required, and a ring network that provides redundancy by connecting L2 (Layer 2) switches in a ring shape has been proposed as a technology that satisfies this requirement. In such a ring network, a technique called ring protection that specifies prevention of a frame loop, path switching at the time of occurrence of a failure, and restoration is an essential requirement. For example, as shown in Non-Patent Document 1, ITU-T Rec. G. 8032 / Y. In 1344, ring protection in Ethernet (registered trademark) called ERP (Ethernet Ring Protection switching) is proposed.

  In this ring protection technology, for a plurality of L2 switches, which are node devices constituting a ring, by controlling the blocking / opening of the ports of each L2 switch, the occurrence of a frame loop failure is prevented and the path is switched.

  In the simplest single link configuration, a frame loop failure is prevented by blocking one port of a certain L2 switch in a normal state. The L2 switch having the blocked port in the normal time is called an RPL (Ring Protection Link) Owner. Each L2 switch learns the frame switching process by creating an FDB (Forwarding Data Base) by referring to the destination MAC address and the source MAC address of the frame flowing in from each port. Thereby, the communication path of the frame from the first L2 switch to the second L2 switch is automatically determined. When a physical link failure occurs between the L2 switches, the port blocking of the RPL Owner is released, and the ports on the failure occurrence side of the L2 switches at both ends of the physical link where the failure occurs are respectively blocked. Initialize the FDB in the L2 switch. Thereby, a detour path at the time of failure is automatically constructed. Although not described here, the ring protection technique has also been proposed for a multi-ring configuration.

Masahiro Maruyoshi, "Ethernet Ring Protection Standardization Trend in ITU-T SG15", IEICE Technical Report, IEICE Technical Report, CS2008-69 (2009-1)

  As described above, when the ring network is used in a wide area network, the L2 switch and the user's network are connected by an access line. In recent years, the speed of the access line has been increasing rapidly. On the wide area network side, in order to reduce the construction cost, the traffic of multiple users is often aggregated. In comparison, there was a case where a communication band could not be secured, resulting in a bottleneck.

  One of the factors is that, in the ring network as described above, the communication path secured at the time of failure is not normally used and there is a vacancy in the communication band, in other words, the accommodation efficiency is low. is there.

  The present invention has been made in view of the above circumstances, and an object thereof is to provide a ring network system having high accommodation efficiency.

  In order to achieve the above object, in the present invention, a ring network including a plurality of L2 switches in which ring protection is implemented and physical links that connect the L2 switches in a ring shape, and each L2 switch is controlled. In a ring network system including a network management device, the network management device is configured such that there is no physical link overlap between the first route and the second route between the first L2 switch and the second L2 switch. And a route calculation means for assigning the first virtual network identifier corresponding to the first route and the second virtual network identifier corresponding to the second route, and from the first L2 switch to the second L2 switch. Each frame or a frame from the second L2 switch to the first L2 switch passes through the first or second path. It comprises setting information distribution means for distributing setting information for each identifier to each L2 switch, and each L2 switch is accommodated in a storage means for storing setting information received from the network management apparatus and another adjacent L2 switch or itself. Switching means for extracting a virtual network identifier from a frame from the access line and performing frame switching processing based on setting information corresponding to the virtual network identifier.

  That is, in the present invention, a plurality of virtual network identifiers for identifying virtual networks are assigned to the same user, and networks with different WAN-side paths are constructed based on the virtual network identifiers. A user can use a plurality of routes at the same time even with a single access line by switching virtual network identifiers according to frames and sending them to a ring network even at the same destination. become. In a ring network, a normal (working) route is assigned to the first virtual network identifier, and a detour route formed when a failure occurs in the normal route is assigned to the second virtual network identifier. . Accordingly, in the present invention, since a plurality of routes can be used simultaneously in the ring network, the storage efficiency is improved.

  As described above, according to the present invention, since a plurality of routes can be used simultaneously in a ring network, storage efficiency is improved.

Network diagram according to the first embodiment Network diagram according to the second embodiment

(First embodiment)
A first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a network diagram according to the first embodiment. The ring network system according to the present embodiment includes a plurality of L2 switches 11 to 14 that are node devices constituting the ring network and trunk lines 31 to 11 that are physical links that connect the L2 switches 11 to 14 in a ring shape. 34, and a network management device 51 that controls and manages the operations of the L2 switches 11-14. The L2 switch 11 accommodates the LAN 80 at the site A via the access line 61. The L2 switch 13 accommodates the LAN 90 at the base B via the access line 62. The communication band (logical band) for each user in the relay lines 31 to 34 constituting the ring is constructed with a smaller band, for example, a half of the physical band of the access lines 61 and 62.

  In this embodiment, each of the L2 switches 11 to 14 conforms to an Ethernet (registered trademark) standard to which a virtual network technology defined by IEEE 802.1Q or the like is applied. The L2 switches 11 to 14 are connected to the ITU-T Rec. G. 8032 / Y. Ethernet (registered trademark) ring protection such as 1344 is applied and implemented.

  In the present invention, a first virtual network ID is assigned to a network that constructs a normal (working) route in a ring network. In addition, regarding the detour route of the switching destination formed by ring protection when it is assumed that the working route (first route) has failed, the detour route (second route) is constructed at the normal time (at the time of working). A second virtual network ID is assigned to the network. It should be noted that both routes do not include overlapping trunk lines. In this embodiment, as described above, Ethernet (registered trademark) VLAN-ID defined by IEEE802.1Q or the like is used.

  The network management device 51 includes a route calculation unit 51a that calculates a first route and a second route between the base A and the base B and assigns a first virtual network ID and a second virtual network ID, respectively. ing. In addition, the network management device 51 includes a setting information distribution unit 51b that distributes setting information for establishing the first route and the second route to the L2 switches 11 to 14 in the ring network. . The setting information is created for each of the first route and the second route, and is identified by the first virtual network ID and the second virtual network ID, respectively.

  Since each L2 switch 11-14 has the same configuration, only the L2 switch will be described in detail here. The L2 switch 11 includes a storage unit 11a that stores setting information received from the network management device 51, and a switching processing unit 11b that performs switching processing based on the setting information stored in the storage unit 11a. The switching processing unit 11b identifies a virtual network ID included in a frame received from any port, and performs switching processing according to setting information corresponding to the virtual network ID.

  Hereinafter, a specific description will be given based on FIG. In the ring network shown in FIG. 1, the first route and the second route using the fact that the route of the frame from the base A to the base B is reversed between the normal time and the failure time due to ring protection. Is calculated. That is, for the frame from the base A to the base B, the clockwise route from the base A to the base B through the path of the L2 switch 11, the L2 switch 12, and the L2 switch 13 is calculated as the first path. If such a route is a normal (working) route, a detour route that becomes a switching destination at the time of failure due to ring protection passes from the base A through the route of the L2 switch 11, the L2 switch 14, and the L2 switch 13. Since this is a counterclockwise route to the base B, the detour route is set as the second route. For the first route, the frame from base A to base B is clockwise, but the frame from base B to base A is counterclockwise. For the second route, base A to base B Note that the frame from to base A is counterclockwise, but the frame from base B to base A is clockwise.

  The route calculation unit 51a of the network management device 51 calculates the first route and the second route, and assigns a different VLAN-ID to each route. In the example of FIG. 1, in a state where no failure has occurred, a VLAN that flows traffic to the first route is assigned as #a, and a VLAN that sends tracks to the second route is assigned as #a.

  The setting information distribution unit 51b of the network management device 51 selects the VLAN− so that the frame transferred from the base A to the base B side in the ring network is selected and the clockwise route is selected for the frame of VLAN-ID # a. For the frame of ID # i, setting information is created for each VLAN-ID for each L2 switch 11-14 constituting the ring so that a counterclockwise route is selected, and distributed to each L2 switch 11-14. In the example of FIG. 1, for VLAN-ID # A, the L2 switch 11 is an RPL Owner, and setting information is created so as to construct a ring network in which the L2 switch 14 side port of the L2 switch 11 is closed. For VLAN-ID # a, the L2 switch 11 is an RPL Owner, and setting information is created so as to construct a ring network in which the L2 switch 12 side port of the L2 switch 11 is closed. It should be noted that in FIG. 1, only the blocked port for VLAN-ID # a is shown and the blocked port for VLAN-ID # a is omitted for the sake of simplicity.

  Each L2 switch 11-14 memorize | stores the setting information for every said VLAN-ID in a memory | storage part, and a switching process part performs a switching process based on the setting information corresponding to every VLAN-ID contained in a flame | frame. Thus, when communication is started, the MAC address of Ethernet (registered trademark) is learned for each VLAN-ID in all L2 switches, and the frame with VLAN-ID # A is sent on the first route on the VLAN. -The frame with ID # a is transferred on the second path.

  When the L2 switches 11 to 14 constituting the ring network detect a failure of the trunk line, the port at the location where the failure has occurred due to the ring protection is blocked, and the port that is blocked in the normal state is opened. The frame is transferred through a communication path that is shifted to either the first path or the second path. In other words, both the frame with VLAN-ID # a and the frame with VLAN-ID # a are transferred to either the first route or the second route. .

  Therefore, the upper limit obtained by adding the flow rates of VLAN-ID #A and #B that flow when the paths are shifted to one another is set in the L2 switches 11 and 13 that accommodate the access lines 61 and 62, respectively. The switching processing units of the L2 switches 11 and 13 perform processing that does not transfer frames exceeding the upper limit in the traffic that sums the VLAN-IDs #A and #A for each output relay line.

  With such a configuration, in normal times, VLAN-ID # a and #a have different output destination relay lines, and each output destination relay line has a traffic upper limit. It is possible to flow the traffic up to the upper limit only with the frame of VLAN-ID # a without being affected by the traffic of the frame of VLAN-ID # i. Similarly, the VLAN-ID # a frame is not affected by the flow rate of the VLAN-ID # a in a normal state, and the VLAN-ID # A is not connected to the trunk line corresponding to the detour path from the first route. You can transfer frames until you reach the upper limit of the total of A and B.

  In the ring network system according to the present invention, the boundary between the access line 61 and the L2 switch 11 or the boundary point between the LAN 80 and the access line 61 at the site A or a terminal in the LAN 80, and the access line 62 and the L2 switch 13. VLAN-ID processing means for assigning or removing VLAN-ID to / from the frame at the boundary point between the LAN 90 and the LAN 90 at the base B and the access line 62 or within the LAN 90 is required. The VLAN-ID processing means may be provided in the L2 switches 11 and 13, for example, may be provided in the line termination device (access device) of the access lines 61 and 62, or in the LANs 80 and 90 which are the main communication entities In other words, it may be built in the terminals in the LANs 80 and 90 that generate frames. The VLAN-ID processing means distributes and assigns VLAN-IDs #A and #A to the frame flowing from the base side to the ring network, and assigns the VLAN-ID from the frame flowing from the ring network. Remove. For the sorting algorithm, for example, when the flow rate of the frame is less than or equal to a predetermined value, either VLAN-ID #a or #a is assigned to all frames, but the flow rate of the frame exceeds the predetermined value Includes a method of alternately assigning VLAN-ID #A or #A to the frame sequence to be transferred. As a result, when communicating with the terminal at the base B from the terminal at the base A, even if a frame is transmitted to the same destination (MAC address), a different VLAN-ID is assigned by the VLAN-ID processing means. , It is possible to flow a frame on both the first route and the second route.

  In addition, when a failure occurs in the relay line and the route is shifted to one side, the route is automatically switched to the detour route in the device constituting the relay line by ring protection. Continuous communication is possible without changing settings such as changing the VLAN-ID assignment rule.

  The ring network system according to the present embodiment has the following advantageous effects. That is, when a ring network is constructed with a bandwidth that is smaller than that of the access line, the relay line becomes a bottleneck in the prior art, and communication from the base A to the base B is performed by one relay line. It becomes communication of the band of minutes. However, according to the present embodiment, if the VLANs for the first or second route are evenly allocated to the frames transmitted from the bases, communication is possible in the band for two relay lines. That is, the accommodation efficiency is improved. However, if the route is misaligned due to a failure of the relay line or the like, it falls to the band of one relay line.

(Second Embodiment)
A first embodiment of the present invention will be described with reference to the drawings. FIG. 2 is a network diagram according to the second embodiment. The first embodiment is an example of one-to-one communication with one ring. On the other hand, this embodiment is an example in the case where the ring is multistage and the number of bases is 3 or more as shown in FIG. The configuration and function of each device are the same as those in the first embodiment.

  In the present embodiment, the route calculation unit 51a of the network management device 51 calculates, as a first route, a route in which a frame from the base A to the base B is clockwise in the entire ring in a multistage ring and turns counterclockwise. Is calculated as the second route, and a different VLAN-ID is assigned to each route. Then, the setting information distribution unit 51b of the network management device 51 sends frames transferred from the base A to the base B side to the L2 switches 11 to 17 constituting the multistage ring, respectively, on the first route or the first route. Create and distribute setting information so that it can be transferred via route 2.

  In the L2 switches 11 to 17 constituting each communication path, the traffic is controlled by a plurality of VLAN-IDs for the first path or the second path instead of individual VLAN-IDs, and the bases A and B side Then, by providing a means for distributing the VLAN-ID, two communications, that is, a first route and a second route that is a detour when the first route is used as an active route and reverses the first route. The route can be used simultaneously.

  In the example of FIG. 2, the first route is selected so that the frame from the base A to the base B is a clockwise route in all rings, and the L2 switch 11, L2 switch 12, L2 from the base A is selected. The data is transferred to the base B through the switch 13 and the L2 switch 14. The second route corresponds to a detour when a failure occurs in the first route, and passes from the base A through the L2 switch 11, the L2 switch 17, the L2 switch 16, the L2 switch 15, and the L2 switch 14. The frame is transferred to the base B. In FIG. 2, for simplification of the drawing, only the blocked port for VLAN-ID # a related to the first route is described, and the blocked port for VLAN-ID # a related to the second route is described. Note that is omitted.

  By the way, if a clockwise and counterclockwise route coexists in a single VLAN-ID route that passes through a plurality of rings, the first route and the second route are switched on a line in which the clockwise and counterclockwise routes are switched in a normal state. Since both of the traffics of the route on the same line flow on the same line, the bandwidth on the line becomes a bottleneck.

  For example, in FIG. 2, regarding the frame from the base A to the base B, if a clockwise route is selected as the first route in the rings #a and #c and only the ring #b is a counterclockwise route, the frame flows in a normal state. The route is transferred from the site A to the site B through the L2 switch 11, L2 switch 12, L2 switch 17, L2 switch 16, L2 switch 13, and L2 switch 14. The detour route for the first route is the route from the base A to the L2 switch 11, L2 switch 17, L2 switch 12, L2 switch 13, L2 switch 16, L2 switch 15, L2 switch 14 and base B. In this case, in the trunk line 38 between the L2 switch 12 and the L2 switch 17 and the trunk line 39 between the L2 switch 13 and the L2 switch 16 indicated by vertical lines in FIG. Overlap. Therefore, in the network management system, for a frame from a certain base to the base, a route that is aligned clockwise or counterclockwise in one VLAN-ID route is calculated.

  By calculating and setting the route in this way, even in the example of FIG. 2, the first route and the second route that is a detour with respect to the first route can be used for two at the same time. Although the speed is one time, the band for two trunk lines can be used at normal times.

  Although the embodiment of the present invention has been described in detail above, the present invention is not limited to this. For example, in each of the above embodiments, the IEEE 802.1Q standard using VLAN-ID is exemplified as a technique for assigning a virtual network ID to a frame, but the present invention can be applied to other standards. In each of the above embodiments, ITU-T Rec. Is used as a ring protection technique in a ring network. G. 8032 / Y. Although the standard of 1344 is illustrated, the present invention can be applied to other standards.

  DESCRIPTION OF SYMBOLS 11-17 ... L2 switch, 11a ... Memory | storage part, 11b ... Switching processing part, 31-39 ... Relay line, 51 ... Network management apparatus, 51a ... Path | route calculation part, 51b ... Setting information distribution part, 61-63 ... Access line , 80, 90, 100 ... LAN

Claims (7)

In a ring network system including a ring network including a plurality of L2 switches on which ring protection is implemented and physical links that connect the L2 switches in a ring shape, and a network management device that controls the L2 switches ,
The network management device calculates the first path and the second path between the first L2 switch and the second L2 switch so that there is no physical link overlap, and the first path corresponding to the first path Route calculation means for assigning the virtual network identifier and the second virtual network identifier corresponding to the second route, a frame from the first L2 switch to the second L2 switch, or the first L2 from the second L2 switch Setting information distribution means for distributing setting information for each virtual network identifier for distributing a frame to the switch to the first or second path to each L2 switch;
Each L2 switch extracts a virtual network identifier from a storage unit that stores setting information received from the network management apparatus, and a frame from an access line accommodated in another adjacent L2 switch or itself, and the virtual network identifier And a switching means for performing a frame switching process based on the setting information corresponding to the ring type network system. When the failure occurs, the switching means of the first L2 switch and the switching means of the second L2 switch have a predetermined upper limit for the total traffic of the frame related to the first virtual network identifier and the frame related to the second virtual network identifier. The ring network system according to claim 1, wherein when it exceeds, a process that does not transfer a frame exceeding the upper limit is performed. The first LAN of the first base is accommodated in the first L2 switch via the access line, and the second LAN of the second base is accommodated in the second L2 switch via the access line. And
The first or second L2 switch that accommodates the access line selects either the first or second virtual network identifier for each frame of the same destination from the access line based on a predetermined rule, The selected first or second virtual network identifier is added to each frame header and transmitted to the ring network, and the first or second virtual network identifier is determined from the frame header of the frame from the ring network to the access line. The ring network system according to claim 1, further comprising a frame processing unit that removes and sends the frame to the first or second LAN. The first LAN of the first base is accommodated in the first L2 switch via the access line, and the second LAN of the second base is accommodated in the second L2 switch via the access line. And
The line terminator arranged at the boundary between the first or second LAN and the access line has a first or second frame for each frame of the same destination sent from the first or second LAN to the ring network. One of the second virtual network identifiers is selected based on a predetermined rule, and the selected first or second virtual network identifier is added to each frame header and transmitted to the ring network. A frame processing means for removing the first or second virtual network identifier from the frame header of the frame for the first or second LAN and sending the frame to the first or second LAN is provided. 2. The ring network system according to 2. The first LAN of the first base is accommodated in the first L2 switch via the access line, and the second LAN of the second base is accommodated in the second L2 switch via the access line. And
The terminal accommodated in the first or second LAN selects either the first or second virtual network identifier based on a predetermined rule for each frame of the same destination sent to the ring network, The ring network system according to claim 1 or 2, further comprising a frame processing unit that adds the selected first or second virtual network identifier to each frame header and transmits the frame header to the ring network. A network management device that controls each L2 switch in a ring network including a plurality of L2 switches in which ring protection is implemented and a physical link that connects the L2 switches in a ring shape,
The first route and the second route between the first L2 switch and the second L2 switch are calculated so that there is no duplication of physical links, and the first virtual network identifier and the first route corresponding to the first route are calculated. Route calculating means for assigning a second virtual network identifier corresponding to the route of 2;
Setting information for each virtual network identifier for the frame from the first L2 switch to the second L2 switch or the frame from the second L2 switch to the first L2 switch to circulate through the first or second route A network management apparatus comprising setting information distribution means for distributing the information to each L2 switch. An L2 switch in a ring network in which a plurality of L2 switches are connected in a ring shape by physical links,
Storage means for storing setting information that can be identified by a virtual network identifier, each of which is a plurality of setting information related to ring protection;
Switching means for extracting a virtual network identifier from a frame from another adjacent L2 switch and performing a frame switching process based on setting information corresponding to the virtual network identifier;
Select either the first or second virtual network identifier based on a predetermined rule for the same destination frame from the access line accommodated by itself, and select the selected first or second virtual network identifier. And a frame processing means for sending to the access line after removing the first or second virtual network identifier from the frame header of the frame to the access line. L2 switch to do.

Images