JP6467237B2 - Path switching device, path switching system, path switching method, and path switching program - Google Patents

Description

  The present invention relates to a path switching device, a path switching system, a path switching method, and a path switching program.

The path redundant configuration is a configuration that enables failure recovery by switching to another path when a failure occurs in a certain path.
Patent Document 1 and Non-Patent Document 1 describe that one path is divided into a plurality of segments (parts), and path switching is performed in units of the segments, thereby improving fault tolerance and effectively using the network. Is described. That is, even in a set of segments constituting the same path, a segment in which a failure has not occurred can be effectively used by distinguishing a segment in which a failure has occurred and a segment in which no failure has occurred.

Japanese Patent Laid-Open No. 11-341209

Qingya She, Xiaodong Huang, and Jason P. Jue, “Survivable Routing for Segment Protection under Multiple Failures”, Optical Fiber Commun. Conf. 2007: Tech.Digest OFC / NFOEC, vol. JThA69, Mar. 2007.

  There may be a case where the communication of the user is interrupted only by the failure recovery means in the segment unit. For example, even if a certain segment is made redundant, there is a severe network failure in which the segment cannot be recovered, such as in the case of a catastrophic disaster. In such a case, a wide range of failure recovery means that bypasses the unrecoverable segment instead of recovering within the segment is required.

  Therefore, the main object of the present invention is to recover a severe failure that cannot be recovered by redundant system switching in a segment unit in a network that is made redundant in a path segment unit.

In order to solve the above problems, a path switching apparatus according to the present invention is a network in which a client path passing through a plurality of segments is set for a segment in which a plurality of communication paths connecting the communication apparatuses form a redundant configuration. A failure information acquisition unit that acquires failure information indicating that the predetermined segment is not communicable due to communication failure between communication devices that are both end points of the predetermined segment,
A switching target path identifying unit that identifies a client path that passes through the predetermined segment that cannot be communicated as a switching target path;
A switching destination path specifying unit that specifies a path that is different from a client path that is common to both ends of the switching target path and that passes through the predetermined segment;
To the a both end points of the switching target path communication device, have a, and a path switching unit for transmitting the instruction to switch to the switching destination path from the switching target path,
The path switching unit backs up information related to the switching target path before switching, and is used in subsequent client path setting processing,
The information related to the switching target path is any of identifier information set for the path, path end point information, QoS (Quality of Service) information, and VPN (Virtual Private Network) information .

As a result, it is possible to recover from a severe failure (both system failures) such that communication between the communication devices at both ends of the segment is impossible. Furthermore, the process of repeatedly setting the same path can be accelerated.

Each of the communication devices of the present invention is
When a part of the segment that is the end point of itself is not communicable, switch to a communicable communication path in that segment,
When a segment whose end point is itself cannot be communicated, the fact is transmitted as failure information to the failure information acquisition unit.

  As a result, each communication device is in charge of a minor failure, and the load is reduced so that the path switching device is in charge of a severe failure, thereby reducing the load on the control device, so that the control device can manage many routers. You can get the scale merit.

The path switching apparatus according to the present invention provides a predetermined failure as a failure that has occurred in a network in which a client path passing through a plurality of segments is set for a segment in which a plurality of communication paths connecting communication apparatuses form a redundant configuration. A failure information acquisition unit for acquiring failure information indicating that the communication device is incapable of communication;
A switching target path identifying unit that identifies a client path that passes through the predetermined communication device that is not capable of communication as a switching target path;
A switching destination path specifying unit for specifying a path different from a client path passing through the predetermined communication apparatus as a switching destination path in common with communication devices that are both end points of the switching target path;
To the a both end points of the switching target path communication device, have a, and a path switching unit for transmitting the instruction to switch to the switching destination path from the switching target path,
The path switching unit backs up information related to the switching target path before switching, and is used in subsequent client path setting processing,
The information related to the switching target path is any of identifier information set for the path, path end point information, QoS (Quality of Service) information, and VPN (Virtual Private Network) information .

As a result, it is possible to recover from a serious failure (device failure) in which one communication device cannot communicate. Furthermore, the process of repeatedly setting the same path can be accelerated.

Each of the communication devices of the present invention switches to a communicable communication path in the segment when a part of the segment that is the end point is not communicable,
The failure information acquisition unit acquires failure information indicating that the predetermined communication device cannot communicate by transmitting a survival confirmation signal to each communication device.

  As a result, each communication device is in charge of a minor failure, and the load is reduced so that the path switching device is in charge of a severe failure, thereby reducing the load on the control device, so that the control device can manage many routers. You can get the scale merit.

  According to the present invention, in a network that is made redundant in units of path segments, it is possible to recover from a severe failure that cannot be recovered by switching the redundant system in units of segments.

FIG. 1A is an explanatory diagram of a client path. FIG. 1B is an explanatory diagram showing a one-side failure of a client path and its countermeasure. FIG. 1C is an explanatory diagram showing various forms of segments that are components of the client path. FIG. 2A is an explanatory diagram when a failure of both systems occurs in a segment through which a client path passes. FIG. 2B is an explanatory diagram when a device failure occurs in a router through which a client path passes. FIG. 2C is an explanatory diagram of a backup path for the client path. It is a block diagram of the control apparatus regarding one Embodiment of this invention. FIG. 4A shows the contents of the database before the failure occurs. FIG. 4B shows the contents of the database after both system failures occur in the segment. FIG. 4C shows the contents of the database after a device failure has occurred in the router. It is a flowchart which shows the troubleshooting processing by the control apparatus regarding one Embodiment of this invention. It is a modification of the flowchart of FIG.

  Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

  FIG. 1A is an explanatory diagram of the client path CP14 set between the routers R1 to R4. Each of the routers R1 to R4 is a packet communication device (transfer device) in the packet communication network. Hereinafter, basic terms used in the present embodiment will be described.

A “segment” is a section of a communication network that connects two routers.
In the following, each segment is described by combining a segment (SEG: Segment) such as “SEG12” and a number indicating the router number at both end points (router R1, R2 if “12”).

A “segment path” is a communication path that connects both end points of a segment.
Hereinafter, as in “P23a”, a segment path is indicated (P: Path), numbers indicating router numbers at both end points of the segment (routers R2 and R3 if “23”), and the operation status in a redundant configuration ( Each segment path is described in combination with an active system “a” illustrated by a solid bold line or a standby system “b” illustrated by a thin broken line.

A “client path” is a logical communication path that connects two routers. A router serving as an end point of a client path is a communication device used by a client such as a base station that accommodates a client terminal.
In the following, each client path is represented by combining a client path (CP: Client Path) such as “CP14” and a number indicating a router number at both end points (routers R1 and R4 if “14”). write.

The “backup path” is a bypass path for avoiding a failure when the client path is the active system. That is, the client path and the backup path have the same two end points (start point, end point), but have different segments from the start point to the end point.
In the following, each spare path is represented by combining a spare path (SP: Spare Path), such as “SP14”, with numbers indicating router numbers at both end points (routers R1, R4 if “14”). write.

  Each path (segment path, client path, backup path) described above is formed on each router as an LSP (Label Switched Path) used in, for example, an MPLS network or an MPLS-TP (Multi Protocol Label Switching-Transport Profile) network. Is done.

First, each of the client path and the backup path passes through one or more segments. For example, the client path CP14 in FIG. 1A sequentially passes through the segments SEG12, SEG23, and SEG34.
Next, the segment may be made redundant or may not be made redundant. For example, the segment SEG34 of FIG. 1A is made redundant by two paths having different paths, a segment path P34a and a segment path P34b.

FIG. 1B is an explanatory diagram showing one-side failure of the client path CP14 and coping with it. Even if a failure occurs in one segment path P34b (hereinafter referred to as a one-system failure), communication can be continued using the other segment path P34a. Similarly, even if a failure occurs in the active segment path P23a, communication can be continued by switching the system to the other standby segment path P23b.
As described above, the system switching in the redundant segment path is autonomously executed when the routers at both end points of the segment detect a failure.

  Therefore, in FIG. 1B, one system failure has occurred in each of the two segments SEG23 and SEG34, but since each is a redundant segment, system switching can be performed in segment path units. That is, the client path CP14 can continue communication between the end point routers R1 and R4 by passing through the segment paths P12a, P23b, and P34a. In other words, fault tolerance is improved by a redundant configuration in segment units.

FIG. 1C is an explanatory diagram showing various forms of segments that are components of the client path CP14. The segment path P23a between the segments SEG23 passes through the router R21 serving as a relay device. Similarly, the segment path P23b passes through the router R22 and the router R23 that are relay devices.
Therefore, since the number of hops is smaller in P23a than in P23b (that is, communication with low delay is possible), it is preferable to adopt the segment path P23a as the active system.
Further, the segment SEG34 in FIG. 1C is not redundant because it is connected only by the segment path P34.

  FIG. 2A is an explanatory diagram when a failure of both systems occurs in the segment SEG23. Although the segment SEG23 is made redundant by two segment paths P23a and P23b, if a failure occurs in both of the segment paths simultaneously (hereinafter referred to as both system failures), communication between the segment SEG23 becomes impossible. . As a result, the communication of the client path CP14 that passes through the segment SEG23 is also interrupted.

  FIG. 2B is an explanatory diagram when a device failure occurs in the router R2. When a failure occurs in the unit of the device called router R2, the segments SEG12 and SEG23 having the device as an end point cannot communicate with each other. As a result, the communication of the client path CP14 passing through the segments SEG12 and SEG23 is also interrupted.

FIG. 2C is an explanatory diagram of the backup path SP14 for the client path CP14 of FIG.
The backup path SP14 and the client path CP14 are common at both end points, but have different relations in their passing paths (passing segments). The client path CP14 passes through the segments SEG12, SEG23, and SEG34. The protection path SP14 passes through the segment SEG14.
Therefore, even if the both-system failure in FIG. 2A or the device failure in FIG. 2B occurs, the protection path SP14 does not pass through the segments SEG12 and SEG23 affected by the failure. That is, the communication between the routers R1 and R4 can be continued by switching to the backup path SP14 instead of the client path CP14.
In this way, switching to the protection path SP14 that bypasses the segment itself is called system switching in units of client paths.

FIG. 3 is a configuration diagram of the control device 1. The control device 1 monitors each router of the packet communication network shown in FIGS. 1 and 2 and controls each router so that communication is continued even if a failure occurs.
The control device 1 is in charge of system switching in units of client paths described in FIG. 2C, but is not in charge of system switching in units of segment paths described in FIG.
The control device 1 and each router are configured as a computer having a CPU (Central Processing Unit), a memory, a hard disk (storage means), and a network interface. The computer executes a program read on the memory by the CPU. Each processing unit is operated.

  The control device 1 includes a failure information receiving unit 11, a switching target path specifying unit 12, a switching destination path specifying unit 13, a path switching unit 14, and a database 15. Details of each component will be described below.

  The failure information receiving unit 11 receives information on both system failures in FIG. 2A and device failures in FIG. The failure information receiving unit 11 may receive a failure detected by each router as an alarm, and whether or not a failure has occurred autonomously from each failure information receiving unit 11 to the router by using a survival confirmation signal or the like. You may inspect.

The switching target path specifying unit 12 refers to the database 15 and specifies a client path that passes (uses) the fault location acquired by the fault information receiving unit 11 as a “switching target path”.
The switching destination path specifying unit 13 refers to the database 15 and can make a detour communication for each switching target path specified by the switching target path specifying unit 12 so as not to pass through the failure part of the switching target path. Such a path is identified as a “switching destination path”. For example, when the client path CP14 in FIG. 2C is a switching target path, the backup path SP14 is the switching destination path.
The path switching unit 14 performs switching control to switch the routers at both end points of the “switching target path” specified by the switching target path specifying unit 12 to the “switching destination path” specified by the switching destination path specifying unit 13. I do.

  The database 15 executes data access to the segment table 15a and the path management table 15b. The segment table 15a manages detailed information for each segment in the packet communication network. The path management table 15b manages detailed information for each client path and backup path in the packet communication network. Details of each table will be described below.

FIG. 4A shows the contents of the database 15 before the failure occurs.
The segment table 15a includes, for each segment (SEG), a router number at each end point thereof, an available bandwidth (segment communication capacity), and a list of client paths or backup paths that pass through the segment (CP / SP list). And correspond.
For example, the segment SEG12 indicated by the record in the first row connects both end points (routers R1, R2), and the bandwidth for T1 can be used. The client path CP14 passes through the segment SEG12.

The path management table 15b associates, for each client path or protection path (CP / SP), a router number at each end point thereof, its use band, and a list of segments (SEG list) through which the path passes.
For example, the client path CP14 indicated by the record in the first line connects both end points (routers R1, R4), and uses the bandwidth for U1. The client path CP14 passes through the segments SEG12, SEG23, and SEG34.

FIG. 4B shows the contents of the database 15 after the failure of both systems has occurred in the segment SEG23.
FIG. 4C shows the contents of the database 15 after a device failure has occurred in the router R2.
4 (b) and 4 (c) are referred to in the description of the flowchart of FIG.

FIG. 5 is a flowchart showing a failure handling process performed by the control device 1.
S <b> 11 is a failure reception determination process by the failure information receiving unit 11. When the failure information reception unit 11 receives from each router that a failure (device failure, both-system failure) has occurred or detects it from polling each router (S11, Yes), the processing proceeds to S12. .

S <b> 12 is a process for specifying a switching target path that passes through the failure location by the switching target path specifying unit 12. The switching target path specifying unit 12 refers to the CP / SP list column of the segment table 15a or the SEG list column of the path management table 15b, and specifies the switching target path that passes through the failure location.
For example, in the case of both system failures in the segment SEG23, the “client paths CP14 and CP24” that are the CP / SP lists in the second row corresponding to the segment SEG23 in the segment table 15a in FIG. Identified.
On the other hand, in the case of a device failure in the router R2, the CP / SP list “client path CP14” of the segment SEG12 corresponding to the end point = R2 and the CP of the segment SEG23 corresponding to the end point = R2 in the segment table 15a of FIG. A union with the / SP list “client paths CP14, CP24” is specified as a switching target path.

  S21 to S25 are loop processes for selecting the switching target paths specified in S12 by the switching destination path specifying unit 13 one by one. For example, since there are two switching target paths “client paths CP14 and CP24” for both the system failure of the segment SEG23 and the device failure of the router R2, loop processing is executed twice (twice).

S <b> 22 is an end point extraction process of the switching target path by the switching destination path specifying unit 13. The switching destination path specifying unit 13 refers to the end point sequence in the path management table 15b and extracts the routers that are the end points of the switching target path.
For example, the first row of the path management table 15b indicates that both end points of the client path CP14 are routers R1 and R4.
Similarly, the third row of the path management table 15b indicates that both end points of the client path CP24 are routers R2 and R4.

S23 is a switching destination path specifying process for sharing an end point by the switching destination path specifying unit 13. The switching destination path specifying unit 13 refers to the end point sequence in the path management table 15b, and specifies the switching destination path that does not pass through the failure point received in S11 while sharing both end points of S22.
For example, the second row of the path management table 15b indicates that there is a backup path SP14 in which the combination of the client path CP14 and the end points (routers R1, R4) matches.
Here, the switching destination path specifying unit 13 may further specify the switching destination path by further referring to information other than the end points. For example, if (used bandwidth U1 of the client path CP14)> (available bandwidth T4 of the segment SEG14 through which the protection path SP14 passes), the capacity of the protection path SP14 may be excluded from the switching destination path.

In S23, there is no path that matches the combination of the two end points (routers R2, R4) of the client path CP24 in any record other than the third row of the path management table 15b. As described above, when the switching destination path does not exist, the switching destination path specifying unit 13 gives up the system switching for each client path and notifies the user by an error message or the like.
Alternatively, both end points may be shared with the client path CP24, and a switchable client path may be newly connected as a switching destination path (after receiving a failure in S11). For example, the switching destination path specifying unit 13 deletes the old client path CP24 (switching target path) in FIG. 4A that has passed through the segments SEG23 and SEG34, and then passes through the segments SEG12 and SEG14 in FIG. 4B. The new client path CP24 (switching destination path) is re-registered.

S24 is a saving process of the switching destination path by the switching destination path specifying unit 13. The switching destination path specifying unit 13 temporarily stores information related to the switching destination path specified in S23 (such as a passing segment) in association with the switching target path.
Similar to S21 to S25, S31 to S35 are loop processes for selecting the switching target paths specified in S12 by the switching destination path specifying unit 13 one by one.

  S32 is a temporary saving process of the switching target path by the path switching unit 14. The path switching unit 14 stores information on the switching target path in a temporary storage table different from the path management table 15b. The information regarding the switching target path to be stored is, for example, identifier information (path number or the like) set for the path, end point information, QoS (Quality of Service) information, VPN (Virtual Private Network) information, or the like.

S33 is an instruction to delete the switching target path by the path switching unit 14. The path switching unit 14 transmits a deletion instruction to the router serving as both end points of the switching target path stored in S <b> 32 and deletes the record indicating the switching target path from the database 15.
Even if the switching target path is deleted from the database 15 in S33, the information related to the path can be reused (restored) when it becomes necessary later by temporarily saving (backup) in S32.

S34 is an instruction to set a switching destination path by the path switching unit 14. The path switching unit 14 sets the switching destination path setting instruction (instruction for operating the backup path by system switching or new connection instruction for the client path) stored in S24 to the switching target path in S33 as its both end points. And a record indicating the switching destination path is newly added to the database 15.
In S33 and S34, since the routers at both end points that are the instruction destinations are the same, the deletion instruction in S33 and the setting instruction in S34 may be included in the same packet, for example.

As described above, as shown in FIG. 4B, the switching destination path (client path CP14 → backup path SP14, client path CP24 before the path change → route) that does not pass through the two system fault locations of the segment SEG23 by the processing of FIG. The system can be switched to the client path CP24) after the change.
Similarly, as shown in FIG. 4C, system switching to a switching destination path (client path CP14 → backup path SP14) that does not pass through the device failure location (segments SEG12, SEG23) of the router R2 can be realized. On the other hand, since the end point of the client path CP24 is originally the router R2, it must be given up switching.

FIG. 6 is a modification of the flowchart of FIG.
First, in FIG. 5, after the processing of the switching destination path specifying unit 13 is performed for each switching target path in the loop of S21 to S25, the path switching is performed for each switching target path in the loop of S31 to S35. Part 14 was done.
On the other hand, in FIG. 6, the loop is integrated into one as S21 to S35, and the processing of the switching destination path specifying unit 13 and the processing of the path switching unit 14 are continuously performed for each switching target path in the loop. This is an example.

As described above, in this embodiment, each router of the packet communication network autonomously recovers a minor failure (single-system failure) for the segment to which it belongs by switching the segment path.
On the other hand, for severe failures such as when a catastrophic disaster occurs (both system failures, router device failures), instead of each router autonomously recovering from the failure, the control device 1 is called a client path or backup path. Responsible for system switching over a wide range across segments.

  In this way, by sharing the role of system switching between a mild failure and a severe failure, the control device 1 can manage many routers by eliminating the need for the control device 1 to handle a minor failure. It is possible to enjoy both the merit of scale and the merit of implementing an early and reliable failure countermeasure (recovery of user communication) because each router does not have to deal with severe failures individually.

1 Control device (path switching device)
11 Failure information receiver (failure information acquisition unit)
12 switching target path specifying unit 13 switching destination path specifying unit 14 path switching unit 15 database 15a segment table 15b path management table

Claims (7)

For a segment in which a plurality of communication paths connecting communication devices form a redundant configuration, communication at both end points of a predetermined segment as a failure that has occurred in a network in which a client path passing through the plurality of segments is set A failure information acquisition unit for acquiring failure information indicating that communication between devices is impossible;
A switching target path identifying unit that identifies a client path that passes through the predetermined segment that cannot be communicated as a switching target path;
A switching destination path specifying unit that specifies a path that is different from a client path that is common to both ends of the switching target path and that passes through the predetermined segment;
To the a both end points of the switching target path communication device, have a, and a path switching unit for transmitting the instruction to switch to the switching destination path from the switching target path,
The path switching unit backs up information related to the switching target path before switching, and is used in subsequent client path setting processing,
The information related to the switching target path is any one of identifier information set for the path, path end point information, QoS (Quality of Service) information, and VPN (Virtual Private Network) information. Switching device. The path switching device according to claim 1 and each of the communication devices are configured.
Each of the communication devices is
When a part of the segment that is the end point of itself is not communicable, switch to a communicable communication path in that segment,
A path switching system characterized in that when a segment that is an end point of itself is incapable of communication, the fact is transmitted as fault information to the fault information acquisition unit. For a segment in which a plurality of communication paths connecting communication devices form a redundant configuration, a predetermined communication device cannot communicate as a failure occurring in a network in which a client path passing through the plurality of segments is set. A fault information acquisition unit that acquires fault information to the effect,
A switching target path identifying unit that identifies a client path that passes through the predetermined communication device that is not capable of communication as a switching target path;
A switching destination path specifying unit for specifying a path different from a client path passing through the predetermined communication apparatus as a switching destination path in common with communication devices that are both end points of the switching target path;
To the a both end points of the switching target path communication device, have a, and a path switching unit for transmitting the instruction to switch to the switching destination path from the switching target path,
The path switching unit backs up information related to the switching target path before switching, and is used in subsequent client path setting processing,
The information related to the switching target path is any one of identifier information set for the path, path end point information, QoS (Quality of Service) information, and VPN (Virtual Private Network) information. Switching device. The path switching device according to claim 3 and each of the communication devices are configured,
Each of the communication devices, when a part of the segment that is the end point of itself is not communicable, is switched to a communicable communication path in the segment,
The path switching system, wherein the failure information acquisition unit acquires failure information indicating that the predetermined communication device cannot communicate by transmitting a survival confirmation signal to each of the communication devices. The path switching device includes a failure information acquisition unit, a switching target path identification unit, a switching destination path identification unit, and a path switching unit,
The failure information acquisition unit, for a segment in which a plurality of communication paths connecting communication devices form a redundant configuration, as a failure occurring in a network in which a client path passing through the plurality of segments is set, Obtain failure information indicating that communication between communication devices at both ends of the segment is not possible,
The switching target path specifying unit specifies a client path that passes through the predetermined segment that cannot be communicated as a switching target path,
The switching destination path specifying unit specifies a path different from a client path that is common to communication devices that are both end points of the switching target path and passes through the predetermined segment as a switching destination path,
The path switching unit transmits an instruction to switch from the switching target path to the switching destination path to a communication device that is both end points of the switching target path ,
The path switching unit backs up information related to the switching target path before switching, and is used in subsequent client path setting processing,
The information related to the switching target path is any one of identifier information set for the path, path end point information, QoS (Quality of Service) information, and VPN (Virtual Private Network) information. Switching method. The path switching device includes a failure information acquisition unit, a switching target path identification unit, a switching destination path identification unit, and a path switching unit,
The failure information acquisition unit, for a segment in which a plurality of communication paths connecting communication devices form a redundant configuration, as a failure occurring in a network in which a client path passing through the plurality of segments is set, Obtain failure information indicating that the communication device cannot communicate,
The switching target path specifying unit specifies a client path that passes through the predetermined communication device that cannot communicate as a switching target path,
The switching destination path specifying unit specifies a path different from a client path that is common to communication devices that are both end points of the switching target path and passes through the predetermined communication device, as a switching destination path,
The path switching unit transmits an instruction to switch from the switching target path to the switching destination path to a communication device that is both end points of the switching target path ,
The path switching unit backs up information related to the switching target path before switching, and is used in subsequent client path setting processing,
The information related to the switching target path is any one of identifier information set for the path, path end point information, QoS (Quality of Service) information, and VPN (Virtual Private Network) information. Switching method. A path switching program for causing the path switching apparatus, which is a computer, to execute the path switching method according to claim 5 or 6 .

Images