JP2012142893A - Network transition method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform fault-tolerant transition in a layer-3 only by switching in a protocol for multiplexing a router and a dynamic routing protocol, in a network in which a plurality of distribution devices are connected to a core layer network and which has no loop configurations.SOLUTION: A network transition method includes: a step of producing a bypass of a layer-2 topology between a transition source network and a transition destination network (S01); a step of changing a server wire connection to a transition destination access switch (S02); a step of changing a sub-system default gateway to a transition destination sub-system distribution device (S03) and exchanging primary-system and sub-system default gateways (S04); a step of making a return communication path accord with an outgoing communication path (S05); a step of changing the sub-system default gateway to a transition destination primary-system distribution device (S06) and exchanging the primary-system and sub-system default gateways (S07); and a step of removing the bypass (S08).

Description

  The present invention relates to a technology for migrating a network configuration, and more particularly to a technology that is effective when applied to a network migration method for non-stop migration from a migration source network to a migration destination network in a network configuration that does not take a loop configuration.

  In recent years, services provided by information processing systems have become longer, and many systems operate continuously for 24 hours 365 days. In particular, in a system that performs mission-critical operations and a system that is constantly used by a large number of users such as a cloud computing environment, an operation that does not allow a short network outage as a service level may be required.

  However, even in a network of such a system, it may be necessary to switch the network in order to shift the network configuration to a new one due to equipment aging or performance problems. At this time, in order to switch the migration source network in operation to the migration destination network, it is necessary to consider appropriate methods and procedures after considering the device specifications and the topology of the migration source / destination networks. There is.

  Here, the network used in the system as described above generally has a redundant configuration in order to increase availability, and this configuration is used to migrate the network while minimizing the impact on system operation. There is a case where the technique of doing is taken.

  As described in Non-Patent Document 1, etc., the network redundancy configuration is largely based on a physical configuration of a loop configuration to make the route redundant, and then STP (Spanning Tree Protocol) or RSTP (Rapid Spanning Tree). Multiple routers and layer 3 (L3) switches using a protocol that multiplexes routers such as VRRP (Virtual Router Redundancy Protocol) and HSRP (Hot Standby Routing Protocol) And the like, and a configuration using a dynamic routing protocol such as OSPF (Open Shortest Path First).

  Among the above network configurations, the configuration using the spanning tree (STP, RSTP) can automatically detect when the network topology changes and reconfigure the tree structure. The operation of shifting by stopping can be performed relatively easily.

  As a technology related to this, for example, Japanese Patent Laid-Open No. 2002-330152 (Patent Document 1) is a method of adding a device to a network by a spanning tree protocol or restarting the operation in the network, When adding or resuming operation, the apparatus is shifted to a state in which only the reception is performed, information in the network is collected in the state in which only the reception is performed, and the existing network topology is not changed by the collected information The network is not reconfigured when a device is added or when a failure occurs, by calculating the priority of the network, and by setting the calculated priority to the own device and then shifting to a transmission / reception enabled state. Describes a technique capable of maintaining network operation.

JP 2002-330152 A

Koichiro Nemoto, “Special Feature: Network Design Standards (Part 1)”, [online], March 12, 2004, ITMedia Corporation, [Search December 2, 2010], Internet <URL: http: // www .atmarkit.co.jp / fnetwork / tokusyuu / jouseki01 / 01.html>

  On the other hand, there are many networks that do not take a loop configuration. For example, in a configuration in which a distribution system such as a plurality of L3 switches is connected as a distribution layer to a core layer network as a general three-layer model network configuration, a configuration in which L3 switches are duplicated without taking a loop configuration In this case, a redundant configuration is adopted by using a protocol such as VRRP for multiplexing routers or a dynamic routing protocol. In such a configuration, for example, when performing migration work such as changing the network topology, it is common to stop the network and provide a maintenance window. It is usually difficult to migrate the network.

  Accordingly, an object of the present invention is to provide a layer 3 network only by switching between a router multiplexing protocol and a dynamic routing protocol in a network configuration in which a plurality of distribution devices are connected to a core layer network and does not take a loop configuration. It is to provide a network migration method for performing non-disruptive network migration at a level. The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

  Of the inventions disclosed in this application, the outline of typical ones will be briefly described as follows.

  A network migration method according to an exemplary embodiment of the present invention includes a case where a plurality of distribution devices are connected to a core layer network, a loop configuration is not used, and a layer 2 topology crosses the distribution devices. A network migration method for performing non-disruptive migration at a layer 3 level from a migration source network to a migration destination network in a network configuration in which a case of crossing access switches of layers is mixed, and includes the following steps: It is what.

  That is, in the network migration method, a layer 2 topology bypass is created between a VLAN segment of the migration source network and the migration destination network to extend the layer 2 topology of the migration source network to the migration destination network. And a second step of switching the connection of the server connected to the access switch of the migration source network to the access switch of the migration destination network.

  Further, after the first step, a protocol for switching the secondary default gateway from the secondary distribution device of the migration source network to the secondary distribution device of the migration destination network and further multiplexing the routers The third step of switching the primary and secondary of the default gateway by the function of and the communication path that is different between the return from the server and the return to the server by the third step, the return to the server A fourth step is executed in which the communication path is changed by the function of the dynamic routing protocol so that the communication path is the same as that from the server.

  Further, after the fourth step, a protocol for switching the secondary default gateway from the primary distribution device of the migration source network to the primary distribution device of the migration destination network, and further multiplexing the routers A fifth step of exchanging the primary and secondary of the default gateway by the function of: and a sixth step of removing a bypass of the layer 2 topology between the migration source network and the migration destination network after the fifth step; Execute.

  Among the inventions disclosed in the present application, effects obtained by typical ones will be briefly described as follows.

  According to a typical embodiment of the present invention, in a network configuration in which a plurality of distribution devices are connected to a core layer network and a loop configuration is not used, a protocol for multiplexing routers and a dynamic routing protocol are used. It is possible to perform non-stop transition of the network at the layer 3 level only by switching.

It is the flowchart which showed the example of the procedure of the network transfer method which is one embodiment of this invention. It is the figure which showed the outline | summary about the example of the network structure before and behind transfer in one embodiment of this invention. It is the figure which showed the outline | summary about the example of the process which produces the bypass of L2 topology in one embodiment of this invention. It is the figure which showed the outline | summary about the example of the process which switches the server connection in one embodiment of this invention. It is the figure which showed the outline | summary about the example of the process which switches the subordinate default gateway in one embodiment of this invention. It is the figure which showed the outline | summary about the example of the process which replaces the primary and secondary default gateway in one embodiment of this invention. It is the figure which showed the outline | summary about the example of the process which changes the return communication path in one embodiment of this invention, and makes it the same communication path as going. It is the figure which showed the outline | summary about the example of the process which switches the subordinate default gateway in one embodiment of this invention. It is the figure which showed the outline | summary about the example of the process which replaces the primary and secondary default gateway in one embodiment of this invention. It is the figure which showed the outline | summary about the example of the process which removes the bypass of L2 topology in one embodiment of this invention. It is the figure which showed the outline | summary about the typical example of the layer 2 topology of the network which does not take a loop structure.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiment, and the repetitive description thereof will be omitted.

<Overview>
A network migration method according to an embodiment of the present invention includes a router such as VRRP in a network configuration in which a plurality of distribution devices such as L3 switches are connected as a distribution layer to a core layer network and a loop configuration is not taken. Network non-disruptive transition at the level of layer 3 is performed using only switching by a redundancy technique such as a protocol for multiplexing network and a dynamic routing protocol. The network configuration here is premised on a configuration in which a layer 2 (L2) topology crosses distribution devices and a case where access layer access switches (such as L2 switches) cross.

  FIG. 11 is a diagram showing an outline of a typical example of a layer 2 topology of a network that does not take a loop configuration. Here, a network segment to which a server device is connected via an access switch (indicated as “ASw” in the figure) is shown, and a solid line between the access switch and the distribution device (indicated as “Dist” in the figure) Indicates that the L2 topology is crossed (the same VLAN (Virtual LAN) is assigned by the distribution device).

  For example, in FIG. 11A, a plurality of primary and secondary access switches are connected to primary and secondary distribution devices to cross the L2 topology, and the primary and secondary distribution devices are connected to cross L2 topology. . In this configuration, for example, when the primary distribution device goes down due to a failure or the like, there is no crossing (connection) of the L2 topology between the primary access switch and the secondary distribution device. The connection to the secondary access switch is switched.

  In FIG. 11B, different segments (thick line and normal line) are configured from the primary and secondary distribution devices. In this configuration, for example, even if the primary distribution device goes down due to a failure, etc., it is possible to reach the secondary distribution device via the transition between access switches. Does not occur. However, since it has a plurality of segments, there are cases where physical placement such as a floor is restricted.

  In FIG. 11C, in the configuration of FIG. 11A, a parent access switch that aggregates a plurality of access switches is provided between the distribution device and the access switch. In this configuration, for example, if either the primary distribution device or the parent access switch goes down due to a failure, etc., the transition between the primary distribution device and the secondary distribution device (or parent access switch) As a result, the connection to the secondary access switch occurs on the server device side.

  In FIG. 11D, in the configuration similar to FIG. 11C, the primary and secondary parent access switches are connected instead of the primary and secondary distribution devices, and the L2 topology is crossed. In this configuration, for example, when the primary parent access switch goes down due to a failure or the like, there is no crossing between the primary access switch and the secondary parent access switch (and distribution device). The connection to the secondary access switch is switched. On the other hand, even if the primary distribution device goes down due to a failure, etc., it is possible to reach the secondary distribution device via the transition between the parent access switches. Does not occur.

  Each topology described above is merely an example, and it is naturally possible to take other topologies. In addition, there may be restrictions on the topology that can be taken due to restrictions on the performance of the equipment used. Note that the network migration in this embodiment is not basically intended for devices that replace devices or change the topology within the segment of the migration source network, but already from the migration source network. The target is to switch to an existing network in another segment. This includes not only switching to a network having a topology different from that of the migration source network, but also switching to a network in which only the device is updated with the same topology. Further, for example, a plurality of migration source networks may be aggregated and integrated into a high performance migration destination network.

  In the present embodiment, a case where the network having the topology shown in FIG. 11A is transferred to the network having the topology shown in FIG. 11D will be described as an example. FIG. 2 is a diagram showing an outline of an example of the network configuration before and after the migration in the present embodiment. In FIG. 2, the upper part shows the network configuration before the migration, and the lower part shows the network configuration after the migration. In addition, each network device in the figure is duplicated by having a primary and a secondary, and the primary device is indicated by a solid line frame and the secondary device is indicated by a dotted line frame.

  The network configuration before the upper transition includes the core 11 (primary) and core 12 (secondary) network devices (switches, etc.) as core layers. In the distribution layer and access layer below the core layer, the left side of the center of the figure shows the old network of the migration source, and the right side shows the new network of the migration destination. That is, in the distribution layer and the access layer, the migration source old network and the migration destination new network coexist.

  In the old network of the migration source, as a configuration corresponding to the topology of FIG. 11A, distribution devices (old Dist 21 (primary), old Dist 22 (secondary)) connected to the core layer devices and access connected to these devices Switch (old ASw41 (primary), old ASw42 (secondary)). The server device 50 is connected to the access switch by the teaming configuration.

  On the other hand, in the new network of the migration destination, as a configuration corresponding to the topology of FIG. 11D, distribution devices (new Dist 23 (primary), old Dist 24 (secondary)) connected to the core layer devices, and these are connected. Parent access switches (parent ASw33 (primary), parent ASw34 (secondary)). The access switch connected to the parent access switch is not yet connected at this point, but a new one may be connected in advance at this point, or the access switch of the old network is connected in the later migration work. You may fix it and continue using it.

  In the old network of the migration source, the segment of the VLAN assigned by the distribution device (the VLAN number is “VLAN12” in the example of FIG. 2), that is, the range where the L2 topology is crossed is indicated by a bold line. In addition, the old Dist 21 is the default system of the default gateway (indicated by a black star), and the old Dist 22 is a sub system (indicated by the white star). In the network configuration before the migration, the communication from the server device 50 is, for example, a route that passes through the old ASw 41, the old Dist 21 (default gateway (correct)), and the core 11 as indicated by a thick arrow in the figure. Return is the reverse path.

  On the other hand, in the network configuration after the migration in the lower stage, the access switch is removed from the old network of the migration source, and in the new network of the migration destination, the access switch (new ASw 43 (correct), new ASw 44 connected to the parent access switch). (Deputy)). In the present embodiment, an example will be described in which the access switch of the new network at the migration destination is replaced with a new one instead of continuously using the one used in the old network at the migration source.

  Further, in this embodiment, it is assumed that the VLAN number assigned to the VLAN by the distribution device in the new network at the migration destination is changed (in the example of FIG. 2, it is changed to “VLAN 34”). It is also possible to move without changing the VLAN number as will be described later. In the network configuration after migration, communication from the server device 50 passes through, for example, the new ASw 43, the parent ASw 33, the new Dist 23 (default gateway (correct)), and the core 11, as indicated by a thick arrow in the figure. The path becomes the reverse path.

  In this embodiment, since a loop configuration is not used as a network configuration, it is not necessary to suppress a loop by using a spanning tree configuration using STP or RSTP. In order to prevent a trouble when an unexpected loop configuration is created due to a change, the definition of STP or RSTP may be set separately.

<Transition procedure>
In the following, a procedure for making a non-stop transition from the network configuration before migration shown in FIG. 2 to the network configuration after migration will be described. FIG. 1 is a flowchart showing an example of a procedure of a network migration method according to an embodiment of the present invention. When the migration operation is started from the state of the network configuration before the migration shown in the upper part of FIG. 2, first, a bypass of the L2 topology is created between the VLAN segments of the migration source old network and the migration destination new network (S01). ). In other words, the broadcast frame in the segment of the old network that is the migration source also flows to the segment of the new network that is the migration destination.

  FIG. 3 is a diagram showing an outline of an example of processing (step S01) for creating a bypass of the L2 topology. In the example of FIG. 3, a bypass of the L2 topology is created by physically connecting the old Dist 21 (primary) and the new Dist 23 (primary) distribution devices. More specifically, the access port (AP) of the old Dist 21 (VLAN number is “VLAN12”) and the access port of the new Dist 23 (VLAN number is “VLAN34”) are physically connected. As a result, the broadcast frame flows from the old Dist 21 to the new Dist 23 regardless of the VLAN number of the access port.

  In this embodiment, since the VLAN number is changed in the new network of the migration destination as an example, it is bypassed to the access port of the new Dist 23. However, if the VLAN number is not changed, the trunk of the new Dist 23 By bypassing to a port, it is also possible to migrate a plurality of other networks (VLANs) all at once to a new network to be migrated to be aggregated and integrated.

  In step S01, the definition of “VLAN34” is added to the trunk port of the distribution device (new Dist 23, 24) and the parent access switch (parent ASw 33, 34) in the new network of the migration destination, thereby further improving the L2 topology. Keep expanding. In this embodiment, assuming that a plurality of other networks with different VLAN numbers are aggregated and integrated into the new network at the migration destination, a definition is added to the trunk port to enable this. However, when there is only one migration source network, the definition may be set for the access port instead of the trunk port.

  Through the above processing, the L2 topology of the migration source old network is extended to the migration destination new network, and as shown by the thick arrows in the figure, the parent ASw 33, the new Dist 23, the old Dist 21 (default gateway), and the core 11 A new route is formed.

  Normally, only one L3 switch can exist in one segment. In this embodiment, the distribution device such as the L3 switch in the new network at the migration destination functions as an L2 switch so that such an L2 topology can be used. Expansion is possible. Therefore, in the new Dist 23 (and the new Dist 24), in order to function as the L2 switch, the function of the L3 switch is stopped. Specifically, an SVI (Switch Virtual Interface) that is a virtual network interface for performing routing at a layer 3 level between VLANs is stopped by an L3 switch command or the like.

  In the present embodiment, the distribution devices of the old Dist 21 and the new Dist 23 are physically connected to create a bypass of the L2 topology. However, the present invention is not limited to this, and the broadcast frame in the segment of the old network of the migration source Can be extended to the segment of the new network of the migration destination, for example, the L2 topology can be expanded by bypassing the old Dist 21 and the L2 switch such as the parent ASw33 of the new network.

  Returning to FIG. 1, next, the access switch is connected to the new network of the migration destination, and the server connection is switched from the access switch of the old network of the migration source to the access switch of the new network of the migration destination (S02). FIG. 4 is a diagram showing an overview of an example of a process of switching server connections (step S02). In the example of FIG. 4, new ASw 43 and 44 are connected as new access switches to the parent ASw 33 and 34 of the new network of the migration destination, respectively, and then the connection of the server device 50 is made in order from the downlink subsystem to the old ASw 41, 42 is switched to the new ASw 42 and 44, respectively. At this time, since the server device 50 is connected to the access switch by the NIC (Network Interface Card) teaming configuration, the connection can be switched without stopping.

  Thereafter, the old ASw 41 and 42 are separated from the old Dist 21 and 22, respectively. As a result of the above processing, the server device 50 communicates through a route passing through the new ASw 43, the parent ASw 33, the new Dist 23, the old Dist 21 (default gateway), and the core 11, as indicated by the thick arrows in the figure.

  As described above, in this embodiment, the access switch of the new network at the migration destination is replaced with the new one from the old network at the migration source. The server device 50 may be reconnected and used continuously. In addition, the processing in step S02, that is, the connection of a new access switch in the new network at the migration destination and the processing for switching the connection of the server device 50 do not need to be performed at this timing, and in step S01 in FIG. After the L2 topology bypass is created, it can be performed at an arbitrary timing.

  Returning to FIG. 1, next, the secondary default gateway is switched from the secondary distribution device of the migration source old network to the secondary distribution device of the migration destination new network (S03). FIG. 5 is a diagram showing an overview of an example of processing (step S03) for switching the secondary default gateway. In the example of FIG. 5, in the old Dist 22, which is the current default gateway of the secondary system, the above SVI is stopped by a command or the like, and the SVI is started by a command or the like in the new Dist 24 to be switched to. As a result, the old Dist 22 functions as an L2 switch, the new Dist 24 functions as an L3 switch, and the secondary default gateway is switched to the new Dist 24.

  Returning to FIG. 1, next, the primary and secondary of the operating default gateway are switched (S04). Specifically, the primary and secondary (active / standby) of the default gateway is switched using a protocol function such as VRRP or HSRP that virtualizes and multiplexes the plurality of router devices described above.

  FIG. 6 is a diagram showing an overview of an example of the process of replacing the primary and secondary default gateways (step S04). In the example of FIG. 6, in the old Dist 21 that is the current primary (active) default gateway, the priority in VRRP or HSRP is lowered from that of the new Dist 24 that is the secondary (standby) default gateway by a command or the like. As a result, the primary gateway of the default gateway is switched, and the new Dist 24 becomes the primary (active) and the old Dist 21 becomes the secondary (standby) default gateway.

  Note that the communication path from the server device 50 to the path through the new ASw 43, the parent ASw 33, the parent ASw 34, the new Dist 24 (default gateway), the new Dist 23, and the core 11 due to the change of the primary and secondary of the default gateway. In other words, the outbound and return communication paths are temporarily different. In this embodiment, in this way, a procedure that enables network transition without interruption is realized by intentionally changing the outgoing and return communication paths temporarily.

  Returning to FIG. 1, next, for the communication path that has become different between going and returning in step S <b> 04, the returning communication path is changed to be the same as the going (reverse to going) communication path (S <b> 05). Specifically, in the dynamic routing protocol such as OSPF, the cost of the interface on the LAN side of the distribution device through which the return communication passes is greatly increased by a command or the like. As a result, the distribution device (primary default gateway) through which the outgoing communication passes becomes lower in cost, and the return communication also passes through this, so the communication path is the same for the outgoing and the outgoing routes.

  FIG. 7 is a diagram showing an outline of an example of processing (step S05) in which the return communication path is changed to be the same communication path as the going. In the example of FIG. 7, the cost of the OSPF of the interface on the LAN side of the new Dist 24 (default gateway) through which the outgoing communication passes is originally high. Here, the cost of the OSPF of the interface on the LAN side of the old Dist 21 that is the distribution device through which the return communication is passed in FIG. 6 is greatly increased to be higher than that of the new Dist 24. As a result, in the return communication due to the OSPF function, the path passes from the core 11 to the new Dist 23, the new Dist 24, the parent ASw 34, the parent ASw 33, and the new ASw 43, and the outbound and return communication paths are the same.

  Returning to FIG. 1, next, the secondary default gateway is switched from the primary distribution device of the migration source old network to the primary distribution device of the migration destination new network (S06). FIG. 8 is a diagram showing an overview of an example of processing (step S06) for switching the secondary default gateway. In the example of FIG. 8, in the old Dist 21 that is the current secondary gateway, the SVI is stopped by a command or the like, and in the new Dist 23 to be switched to, the SVI is started by a command or the like. As a result, the old Dist 21 functions as an L2 switch, the new Dist 23 functions as an L3 switch, and the secondary default gateway is switched to the new Dist 23.

  Returning to FIG. 1, next, the primary and secondary of the operating default gateway are switched (S07). Specifically, as in step S03, the primary / secondary (active / standby) of the default gateway is switched using a function of a protocol such as VRRP or HSRP.

  FIG. 9 is a diagram showing an overview of an example of the process of replacing the primary and secondary default gateways (step S07). In the example of FIG. 9, in the new Dist 24 that is the current primary (active) default gateway, the priority in VRRP or HSRP is lowered from that of the new Dist 23 that is the secondary (standby) default gateway by a command or the like. As a result, the primary gateway of the default gateway is switched, and the new Dist 23 becomes the primary gateway (active), and the new Dist 24 becomes the secondary gateway (standby) default gateway. As the default gateway is switched, the communication from the server device 50 is changed to a route passing through the new ASw 43, the parent ASw 33, the new Dist 23 (default gateway), and the core 11, and the return is the reverse route. .

  Returning to FIG. 1, the L2 topology bypass created in step S01 that is finally unnecessary is removed (S08), and the series of procedures is terminated. FIG. 10 is a diagram showing an overview of an example of the process (step S08) for removing the bypass of the L2 topology. As shown in FIG. 10, by removing the physical connection for creating a bypass between the old Dist 21 and the new Dist 23, the same network configuration as that after the migration shown in FIG. 2 is obtained.

  Note that, in the network migration method including the above-described series of procedures, each procedure in the flowchart of FIG. 1 may be performed manually, or the processes in steps S03 to S07 that do not involve physical work such as connection change. Can be implemented by a script or program and automatically executed by each distribution device. For example, the physical connection work in steps S01 and S02 is performed in advance, and the processes in steps S03 to S07 and their operation confirmation processes are performed based on instructions from an operation management system or the like that manages and monitors the network. The network migration work can be semi-automated by automatically executing the programs, etc., respectively mounted in cooperation with each other at an appropriate timing.

  By semi-automating a series of procedures, it is possible to efficiently perform network migration in a short time. On the other hand, when performing a series of manual procedures sequentially, in this embodiment, the network is stopped through each procedure. Therefore, it is possible to incorporate a more detailed operation check and confirmation of network configuration change / reflection status each time after execution of each procedure, enabling more reliable network migration. It becomes.

  For example, in this embodiment, as described above, the communication path is changed by manipulating parameters in protocols such as VRRP and HSRP and costs in dynamic routing protocols such as OSPF. It is possible to confirm whether or not the configuration has been properly changed or reflected, and whether or not communication has really stopped. For example, when the cost in OSPF is changed, the routing table is updated also in other peripheral devices, so that it can be confirmed including these. Even if the routing table is not updated correctly for some reason, it is possible to determine whether or not to fall back each time a procedure is executed.

  As described above, according to the network migration method according to an embodiment of the present invention, in a network configuration in which a plurality of distribution devices such as L3 switches are connected to the core layer network and a loop configuration is not taken. The L2 topology is expanded by causing the L3 switch of the migration destination network to function as the L2 switch and creating a bypass between the migration source network and the migration destination network. Furthermore, the network configuration and the communication route are sequentially shifted while temporarily changing the going and returning communication paths only by switching by the redundancy technology of the router multiplexing protocol and the dynamic routing protocol. As a result, it is possible to make a non-stop transition of the network at the layer 3 level without temporarily taking a loop configuration.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiment. However, the present invention is not limited to the embodiment, and various modifications can be made without departing from the scope of the invention. Needless to say.

  For example, in the present embodiment, a description is given assuming a transition to a new network associated with replacement of devices, that is, a network transition associated with maintenance work. A utilization method is also possible in which a network to be used at any time among a plurality of networks is switched without interruption at a layer 3 level according to business requirements. For example, it is possible to use such as switching communication between networks composed of distribution devices having different security function specifications depending on the time zone, and returning to the original state.

  INDUSTRIAL APPLICABILITY The present invention can be used in a network migration method that performs non-stop migration from a migration source network to a migration destination network in a network configuration that does not take a loop configuration.

11 ... core (positive), 12 ... core (secondary),
21 ... Old distribution device (Dist) (primary), 22 ... Old Dist (secondary), 23 ... New Dist (primary), 24 ... New Dist (secondary),
33 ... Parent access switch (ASw) (primary), 34 ... Parent ASw (secondary),
41 ... Old ASw (primary), 42 ... Old ASw (secondary), 43 ... New ASw (primary), 44 ... New ASw (secondary),
50: Server equipment.

Claims (4)

In a network configuration in which a plurality of distribution devices are connected to the network of the core layer, a loop configuration is not taken, and a case where the layer 2 topology crosses the distribution device and a case of crossing the access switch of the access layer are mixed. A network migration method for performing non-disruptive migration at a layer 3 level from a migration source network to a migration destination network,
A first step of creating a layer 2 topology bypass between the source network and the destination network VLAN segment to extend the source network layer 2 topology to the destination network;
A second step of switching the connection of the server connected to the access switch of the migration source network to the access switch of the migration destination network;
Protocol function for switching the secondary default gateway from the secondary distribution device of the migration source network to the secondary distribution device of the migration destination network after the first step, and further multiplexing the router A third step of swapping the primary and secondary of the default gateway by:
For the communication path that is different between the return from the server and the return to the server by the third step, the return communication path to the server is changed by the function of the dynamic routing protocol, and the server A fourth step with the same communication path as from
Protocol function for switching the secondary default gateway from the primary distribution device of the migration source network to the primary distribution device of the migration destination network after the fourth step, and further multiplexing the routers A fifth step of replacing the primary and secondary of the default gateway by:
After the fifth step, a network transition method comprising: performing a sixth step of removing a bypass of the layer 2 topology between the migration source network and the migration destination network. The network migration method according to claim 1,
When creating the bypass of the layer 2 topology between the migration source network and the migration destination network in the first step, the distribution device of the migration destination network is allowed to function as a layer 2 switch. A featured network migration method. In the network migration method according to claim 1 or 2,
When creating the bypass of the layer 2 topology between the migration source network and the migration destination network in the first step, the access to which the first VLAN number of the distribution device of the migration source network is assigned A network migration method comprising: connecting a port to an access port to which a second VLAN number of a distribution device of the migration destination network is assigned. In the network migration method according to claim 1 or 2,
When creating the bypass of the layer 2 topology between the migration source network and the migration destination network in the first step, the access to which the first VLAN number of the distribution device of the migration source network is assigned By connecting a port and a trunk port to which the first VLAN number of the distribution device of the migration destination network is assigned, a non-disruptive migration is performed from a plurality of migration source networks to the migration destination network all at once. A network migration method characterized by the above.