JP6774350B2 - Optical node system, ring network system, and disaster recovery method - Google Patents

Description

The present invention relates to an optical node system , a ring-shaped network system, and a technique for a failure recovery method.

The two indicators required for a network, long transmission distance and improved reliability during transmission, are trade-offs with each other.
First, the "long transmission distance" will be described. A packet signal transmitted from a transmitter in the form of an electric signal or an optical signal is attenuated according to a propagation distance, so that the transmission distance has a physical limit. As a guideline for the transmission distance, an optical fiber line can reach up to several tens of kilometers by a non-relay optical transmission method. That is, when it is desired to transmit a signal between points of several hundred kilometers, it is necessary to provide an optical node device as a relay point every several tens of kilometers to amplify the attenuated signal.

Next, "improvement of reliability during transmission" will be described. In a network configuration in which optical node devices at relay points are connected in series, if a device failure occurs somewhere, the entire route in series will be interrupted. That is, even if the failure probability of each node is low, the reliability of the entire route becomes low as the number of nodes increases.

Therefore, by constructing a network in which series routes are connected in a ring shape and adopting a redundant configuration that uses the reverse route in the event of a one-way route failure, the failed node can be bypassed. Can be done.
FIG. 8 is a ring-shaped network system. In this network system, a plurality of geographically separated base groups (bases A, B, and C) are connected in a ring shape by lines a, b, and c that connect the bases with a transmission medium such as an optical fiber. There are two communication paths between bases, right system (clockwise) and left system (counterclockwise), and if a failure occurs on one route, the other route is redundant. By using it as a route, reliability is ensured. That is, each line a, b, c accommodates communication for two systems.

FIG. 9 is a configuration diagram showing details of optical node devices deployed at each base of the network system of FIG. The node A deployed at the base A in FIG. 8, the node B deployed at the base B, and the node C deployed at the base C are light such as a switch node of MPLS-TP (Non-Patent Document 1), respectively. It is a node device (FIG. 3, PTN node of Non-Patent Document 2 corresponds to this).
Each optical node device has a plurality of optical I / F units (first IF11, second IF12, third IF13), and a packet processing unit 10 for relaying communication between the optical I / F units. ..

The packet processing unit 10 processes a packet configured as an electric signal.
The optical I / F unit converts an electric signal packet received from the packet processing unit 10 into an optical signal and transmits it to the outside by a transmitter (TX). Further, the optical I / F unit converts an optical signal received by the receiver (RX) from the outside into an electric signal and notifies the packet processing unit 10. The optical I / F section is modularized as a set of a combination of TX and RX, and is mounted in accordance with a common standard (SFF-8431, Non-Patent Document 3) such as SFP (Small Form-factor Pluggable).
An optical node device (optical I / F unit, packet processing unit 10) that performs electrical operation generally has a higher failure rate than an optical fiber line that only operates passively with respect to light. In this paper, it is assumed that the failure rate of parts such as optical fiber lines is sufficiently low.

Here, even in a redundant configuration using a ring-shaped network system, it is not possible to deal with simultaneous failures at two or more locations. Therefore, in addition to making the entire route redundant, which is a ring-shaped network, a form is used in which redundancy is also performed for each device. That is, one optical bypass device is added to each optical node device constituting the path to form a detour (Non-Patent Document 4).
In the ring-shaped network system, not only the user packet but also the S (Section) -OAM packet and P for fault detection specified in Non-Patent Document 5 and the like as MPLS-TP OAM (Operations, Administration, and Maintenance). (Path) -OAM packets are mixed.
Also, as defined in Non-Patent Documents 6 and 7, OAM packets are also encapsulated in Ethernet®.

FIG. 10 shows a configuration in which an optical node device and an optical bypass device are combined. Optical bypass switches 11SS and 12SS are provided as optical bypass devices between the lines a to c and the optical node devices (nodes A to C).
When the optical node device is "normal", the optical bypass device connects the lines a to c with the optical node device by conducting the upper port and the lower port with each other at an optical level. This optical bypass device is an optical component that performs only passive operation, and the failure rate is as low as that of an optical fiber line.

When each optical node device detects its own abnormality due to a device failure or power failure due to a power failure, the two optical bypass devices connected to it are switched to the "abnormality" mode. In this mode, the optical bypass device keeps the upper port and the lower port folded inside to prevent signals from being mixed into the optical node device when a failure occurs.
For example, when the node A is abnormal, the right packet (broken arrow) from the line a to the node A is not sent to the node A but is sent to the opposite line c by the optical bypass switch 12SS immediately before the node A. Bypass. That is, the optical bypass switch 12SS diverts the signal directed to the receiver (RX) of the second IF12 to the signal line transmitted by the transmitter (TX) of the first IF11.
Similarly, when node A is abnormal, the left packet (solid arrow) from line c to node A is not sent to node A by the optical bypass switch 11SS immediately before it, but to the opposite line c. Detour to let it flow. That is, the optical bypass switch 11SS diverts the signal directed to the receiver (RX) of the first IF 11 to the signal line transmitted by the transmitter (TX) of the second IF 12.

Due to the trade-off relationship described above, there is a limit to the transmission distance that the optical node device does not relay. Therefore, it is desirable that the distance of each line between the optical node devices is shorter (with a margin) than the limit value of the transmission distance in consideration of the case where a plurality of lines are connected by the optical bypass device.
In this way, by making each optical node device redundant individually by the optical bypass device, even if a failure occurs in three or more optical node devices at the same time, the optical bypass device recovers individually at each failure location ( Bypassing) is possible, and reliability is improved.

Murakami, et al., "Trends in International Standardization of MPLS-TP", [online], March 2013, [Search on February 7, 2017], Internet <URL: http://www.ntt.co.jp/ journal / 1303 / files / jn201303053.pdf> Kawasaki, et al., "Practical development of 100Gbits high-speed, large-capacity packet transport system", [online], 2016, [Search on February 7, 2017], Internet <URL: http://search.ieice.org /bin/pdf_link.php?category=B&fname=j99-b_4_275&lang=1year=2016> SFF Committee, "SFF-8431 (SFP / SFP +) Pluggable Module Specifications", [online], July 2009, [Search February 7, 2017], Internet <URL: ftp://ftp.seagate.com /sff/SFF-8431.PDF〉 Mitsubishi Cable Industries, "Optical Bypass Switch", [online], October 2008, [Search on February 7, 2017], Internet <URL: http://www.mitsubishi-cable.co.jp/jihou/ pdf / 105 / t01.pdf> Information and Communication Technical Committee, "OAM mechanism of MPLS-TP in packet transport network (PTN)", [online], February 2014, [Search on February 7, 2017], Internet <URL: http: / /www.ttc.or.jp/jp/document_list/pdf/j/STD/JT-G8113.1v1.pdf> IEEE, "IEEE802.3ah Standard", 57. Operations, Administration, and Maintenance (OAM), 2004 ITU-T, "Y.1730 Recommendation", [online], May 2006, [Search February 7, 2017], Internet <URL: https://www.itu.int/rec/T-REC -Y.1730 / en>

The configuration in which the optical bypass device and the optical node device as illustrated in FIG. 10 are used in combination as separate devices includes a signal line connecting the optical bypass device and the optical node device, a power supply line to the optical bypass device, and the like. Network configuration will increase.
For example, the optical bypass device of FIG. 10 is connected across two optical I / F units housed in the same optical node device, and is connected to the optical I / F unit on the transmitter (TX) side of the connection destination. The optical I / F section on the receiver (RX) side is separate. In such a connection configuration that spans the two optical I / F units, the distance between the two optical I / F units (ports) may deviate from the SFP + standard.

Therefore, the main subject of the present invention is to realize high reliability at the time of transmission with a small burden for a network that realizes a long transmission distance.

In order to solve the above problems, the optical node system of the present invention has the following features.
The present invention is an optical node system for an optical node equipment is a plurality mountable optical I / F unit for connecting optical fibers,
The optical node device includes a packet processing unit that receives a first optical I / F unit or a second optical I / F unit as the wearable optical I / F unit.
The optical node system includes an optical bypass device that forms a bypass path outside the optical node device when the second optical I / F unit is mounted on the optical node device.
The first optical I / F section
A receiver that connects the optical fiber for reception and
A transmitter for connecting the optical fiber for transmission and
It has an optical bypass mechanism configured to straddle a reception signal line input from the optical fiber to the receiver and a transmission signal line output from the transmitter toward the optical fiber.
The second light I / F section
A receiver that connects the optical fiber for reception and
It has a transmitter for connecting the optical fiber for transmission, and has a transmitter.
The packet processing unit
When the first optical I / F unit is attached, the reception signal line is normally conducted to the receiver and the transmission signal line is conducted to the optical bypass mechanism in the first optical I / F unit. Is conducted to the transmitter, and in the event of an abnormality, a bypass path is formed so as to fold back a signal from the receiving signal line to the transmitting signal line.
When the second optical I / F unit is attached, the optical bypass device is characterized in that a bypass path is formed in the optical bypass device so as to bypass the optical fiber for reception to the optical fiber for transmission in the event of an abnormality.

As a result, the line bypassed by the optical bypass mechanism is closed inside one optical I / F section, so that the individual optical I / F sections can be freely used without being affected by the physical layout of the optical node device. It can be made removable. Therefore, for a network that realizes a long transmission distance, high reliability at the time of transmission can be realized with a small burden.
Further, in the present invention, the packet processing unit is configured as a config circuit capable of changing the internal wiring according to the config setting.
An accelerator card including the config circuit and the attachment / detachment portion of the optical I / F unit is connected to the optical node device.
The bypass path according to the config setting different between the case where the first optical I / F portion is attached to the detachable portion and the case where the second optical I / F portion is attached to the detachable portion. It is characterized by changing the internal wiring for forming the.

In the present invention, the optical node device further includes a control unit that controls the optical bypass mechanism during normal and abnormal conditions.
When the control unit fails in a predetermined first optical I / F unit among the plurality of first optical I / F units, the optical bypass inside the predetermined first optical I / F unit is used. It is characterized in that the mechanism is instructed to form the bypass path, and the other first optical I / F unit is made to transmit a packet for notifying the occurrence of its own failure to another device. And.

As a result, even when the first optical I / F section having the optical bypass mechanism fails by itself, a bypass path is formed only in the first optical I / F section, so that reliability can be ensured by redundancy. While obtaining it, it is not necessary to affect other first optical I / F sections.

In the present invention, when the control unit fails in the packet processing unit inside the optical node device for relaying communication between a plurality of the first optical I / F units, the control unit owns the optical node device. It is characterized in that each of the plurality of first optical I / F portions mounted on the first optical I / F portion is instructed to form the bypass path in the optical bypass mechanism inside the first optical I / F portion. ..

As a result, in response to a device failure of the optical node device, by forming a bypass path collectively for a plurality of first optical I / F units of the optical node device, redundancy can be realized for each device.

The present invention forms a ring-shaped network of the optical fibers in which the set of the optical node systems passes from the predetermined optical node system through another optical node system and returns to the predetermined optical node system. The lines between the two optical node systems constituting the set of the optical node systems are redundantly connected by two systems having different signal transmission directions from each other.
Each of the first optical I / F units has the receiver for connecting the optical fiber of one of the two systems and the transmitter for connecting the optical fiber of the one system. It is characterized by being.

As a result, by using both the redundancy in the end-to-end path by the ring-shaped network and the local redundancy by the optical bypass mechanism in the first optical I / F part of the optical node device, it is possible to perform multiple locations. It will be possible to deal with difficult cases such as simultaneous failures.

According to the present invention, it is possible to realize high reliability at the time of transmission with a small burden on a network that realizes a long transmission distance.

It is a ring-shaped network system according to this embodiment. The details of the optical node device of FIG. 9 according to this embodiment are shown. It is the detail of the optical node apparatus which comprises the ring-shaped network system which applied the present invention of FIG. 1 which concerns on this Embodiment. It is a hardware block diagram of the ring-shaped network system of FIG. 1 which concerns on this embodiment. It is a circuit configuration on the accelerator card of FIG. 4 according to this embodiment. It is a flowchart which shows the process about P-OAM packet which concerns on this Embodiment. FIG. 7A shows a case where one optical I / F unit transmits and receives one system. FIG. 7B shows a case where two systems are transmitted and received by one optical I / F unit. It is a ring-shaped network system. It is a block diagram which shows the detail of the optical node apparatus deployed in each base of the network system of FIG. It is a configuration that combines an optical node device and an optical bypass device.

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

FIG. 1 is a ring-shaped network system to which the present invention is applied. Compared with the ring-shaped network system of FIG. 10, the optical bypass device added as a device separate from the optical node device in FIG. 10 and the optical bypass added inside the optical I / F section of the optical node device in FIG. 1 It has been replaced by a mechanism (bypass switch portions 11S, 12S).
The optical bypass mechanism bypasses inside one optical I / F section in which it is housed.
The bypass switch unit 11S in the first IF11 bypasses the signal directed to the receiver (RX) of the first IF11 to the signal line transmitted by the transmitter (TX) of the first IF11.
The bypass switch unit 12S in the second IF12 bypasses the signal directed to the receiver (RX) of the second IF12 to the signal line transmitted by the transmitter (TX) of the second IF12.

FIG. 2 shows the details of the optical node device of FIG. On the other hand, the details of the optical node device of FIG. 1 will be described later in FIG.
The optical node device has a plurality of optical I / F units (first IF11, second IF12, third IF13) and a packet processing unit 10.
The first port control circuit 11C controls the first IF 11 mounted on the first port 11P.
The second port control circuit 12C controls the second IF 12 mounted on the second port 12P.
The controller 19C detects the failure of the packet processing unit 10 and enables the detour of the optical bypass device.

The packet processing unit 10 includes a label switch unit 20, S-OAM removal units 31, 33, S-OAM mixing units 32, 34, label removal units 41, 43, labeling units 42, 44, and a redundant route. It has a selection unit 51 and a redundant route creation unit 52.
The label switch unit 20 identifies the label of the received packet and forwards it to the output port specified in the routing table. Therefore, the label switch unit 20 has 1 input 2 output distribution units 21, 23, 26, 28 and 2 input 1 output confluence units 22, 24, 25, 27.
Here, the inside of the label switch unit 20 is divided into four ports (Port A, B, C, D), and one distribution unit and one merging unit correspond to one port.
That is, the label switch unit 20 includes Port D having a distribution unit 21 and a merging unit 22, Port C having a distribution unit 23 and a merging unit 24, and Port B having a merging unit 25 and a merging unit 26. It has a Port A having a confluence 27 and a distribution 28.

The merging section 27 of Port A includes one external output line (input line of the label removing section 43) and two input lines (output line of the distribution section 21 of Port D facing Port A (not shown). It is connected to the output line of the distribution unit 23 of Port C facing Port A).
Further, the distribution unit 28 of Port A includes one external input line (output line of the labeling unit 44) and two output lines (input line of the merging unit 22 of Port D facing Port A (not shown). It is connected to the input line of the confluence 24 of Port C facing Port A).
In this way, one port is connected to two opposite ports.

First, the transfer process of the user packet when the optical node device is operating normally will be described.
The user packet is transmitted along the route of "transmission node-> relay node-> destination node". Here, by assigning a unique MPLS-TP label to each user packet, it is possible to independently provide reachability to the destination node while sharing the same network among a plurality of users.
The user packet input from the third IF 13 is duplicated and notified to the left-based labeling section 42 and the right-based labeling section 44 via the redundant route creation section 52, respectively. Hereinafter, the processing content of the user packet will be described focusing on the right system, but the same applies to the left system. The label-giving unit 44 assigns a specified MPLS-TP label to the notified user packet, and then notifies the label switch unit 20 (distribution unit 28).

Here, the label switch unit 20 determines the transfer destination port of the user packet depending on which device on the route of the user packet the optical node device itself is.
(1) When its own optical node device is a transmitting node, the label switch unit 20 transfers the right-type user packet to Port C (first IF11) and the left-type user packet to Port D (second IF12). ..
(2) When its own optical node device is a relay node, the label switch unit 20 transfers the left user packet from Port C to Port D so as to transfer the right user packet from Port D to Port C. Forward.
(3) When its own optical node device is the destination node, the label switch unit 20 transfers the right-type user packet to Port A and the left-type user packet to Port B.

Consider the case where the own optical node device transmits a user packet to the next hop optical node device as (1) transmitting node or (2) relay node. At this time, the user packet from the label switch unit 20 passes through the S-OAM mixing units 32 and 34 and is notified to the optical I / F unit (first IF11, second IF12) of the transmission destination. The optical I / F unit transmits the notified user packet to the next-hop optical node device heading for the destination.

Consider the case where the own optical node device receives a user packet from the previous hop optical node device as (2) relay node or (3) destination node. At this time, the user packet from the optical I / F unit passes through the S-OAM removal units 31 and 33 and is notified to the label switch unit 20. (2) In the case of the relay node, the label switch unit 20 transmits the notified user packet to the optical I / F unit (second IF12) different from the received optical I / F unit (first IF11). .. (3) In the case of the destination node, the label switch unit 20 notifies the notified user packet to the optical I / F unit (third IF13) for the destination.

When the user packet heading from the label switch unit 20 to the third IF 13 passes through the label removing units 41 and 43, the MPLS-TP label attached to the labeling units 42 and 44 is removed.
The redundant route selection unit 51 determines a user packet to be output to the third IF 13 based on the user packets of both the left and right systems that have passed through the label removal units 41 and 43. For example, the redundant route selection unit 51 outputs the right user packet when the user packets of both systems can be received normally, and the normal user packet when only the user packet of one system can be received normally. Is output.

The user packet forwarding process has been described above. The optical node device handles not only user packets but also control packets (S-OAM packets, P-OAM packets) for realizing MPLS-TP OAM functions such as failure notification.
The S-OAM mixing units 32 and 34 cause the user packets transmitted from the optical I / F unit to be mixed with the S-OAM packets in the transmission path at a constant ratio and transmitted. This S-OAM packet contains information for notifying other devices of the failure status of the own device. The S-OAM removing units 31 and 33 remove the mixed S-OAM packets to prevent the S-OAM packets from leaking to the outside as user packets.
By sending and receiving S-OAM packets between optical node devices, it is possible to help identify the location of a failure. For example, when the user packet passing through the route of "node A-> line a-> node B" is interrupted, it is considered that the cause is a failure of node A, a failure of node B, a disconnection of line a, or the like.

For example, if the transmitting node A can transmit an S-OAM packet flagged with an RDI signal indicating the occurrence of its own failure to the node B, it can be seen that the cause of the user packet interruption is the node A failure. In this case, only the failed optical I / F section in the node A may be bypassed by the optical bypass device.

Alternatively, if node B on the receiving side can send an S-OAM packet flagged with an RDI signal due to non-delivery of the S-OAM packet to node A, the cause of the user packet interruption is the receiving side light of node B. It turns out that it is an obstacle of the I / F part. In this case, only the optical I / F portion in the node B may be bypassed by the optical bypass device.
That is, by using the S-OAM packet, it is possible to minimize the location where the bypass processing by the optical bypass device is activated, and it is possible to reduce the influence of the ring network failure countermeasure.
The P-OAM packet is also a control packet for the OAM function similar to the S-OAM packet, and the details thereof will be described later.

Next, the details of various control signals when a failure occurs in the optical node device will be described. The control signals shown are an RDI (Remote Defect Indication) signal, a LOC (Loss of Connection) signal, a LOS (Loss of Signal) signal, an ART signal, an SA signal, an SI1 signal, and an SI2 signal. Hereinafter, the range of failure occurrence will be described by dividing it into two cases.
-Case 1 (IF failure): When a failure occurs only in a specific optical I / F section, but the other optical I / F section and the packet processing section 10 are normal. Here, SI1 signal and SI2 signal are used for failure notification.
-Case 2 (device failure): When a failure occurs in the entire optical node device due to a power failure or the like, and it is necessary to bypass the optical node device itself. Here, an SA signal is used for failure notification.

Case 1 (IF failure) will be described.
The first port control circuit 11C indicates that the LOS signal from the first IF 11 of the first port 11P to be connected or the S-OAM packet connected to the first IF 11 is not detected for a certain period of time from the S-OAM removing unit 33. Receives a LOC signal. As a result, it is possible to recognize that a failure has occurred in the reception processing of the first port 11P.
Then, the first port control circuit 11C that recognizes the failure transmits the RDI signal to the S-OAM mixing unit 34. The S-OAM mixing unit 34 notifies the opposite optical node device of the failure by setting an RDI signal flag in the S-OAM packet to be mixed by the S-OAM mixing unit 34.
The above is the details of the control signal processing between the first port control circuit 11C and the first port 11P, but the same applies to the second port control circuit 12C and the second port 12P.

Further, in case 1, it is necessary to switch the optical bypass device to the state at the time of abnormality. Therefore, the first port control circuit 11C at the time of failure transmits the SI1 signal to the optical bypass switch 11SS. Similarly, the second port control circuit 12C at the time of failure transmits the SI2 signal to the optical bypass switch 12SS.
Upon receiving these SI1 signal and SI2 signal, the optical bypass device changes from a normal conduction state to an abnormal return state. In other words, the SI1 signal and the SI2 signal are not transmitted under normal conditions.
Further, after the SI1 signal is transmitted, the user packet does not arrive at the first IF 11 due to the return processing of the optical bypass device. Therefore, the first port control circuit 11C also transmits the SI1 signal to the first IF11 to shut down the power of the first IF11.

Case 2 (device failure) will be described. This is a case where a user packet whose starting node, relay node, or ending node is a certain optical node device is interrupted due to a device failure of the optical node device.
When a failure occurs in each processing unit (particularly, the label switch unit 20 which is the main unit) in the packet processing unit 10, the controller 19C detects that fact by an ART signal or the like. Alternatively, the controller 19C may detect a hang-up by setting a watchdog timer in each processing unit in the packet processing unit 10 and periodically acquiring a state from each processing unit.

Then, the controller 19C controls to bypass the optical node device itself without using it as a passage path for the user packet. Specifically, the controller 19C changes both of the two optical bypass devices to the "return state at the time of abnormality" by transmitting the SA signal to the two optical bypass devices (optical bypass switches 11SS and 12SS). As a result, a ring network structure in which the own optical node device is completely excluded is formed, so that it is possible to prevent the spread of obstacles to other nodes. For example, the route of "node A-> node B-> node C" is switched to the route of "node A-> (optical bypass mechanism of node B)-> node C" due to a device failure of node B. As a result, communication is continued without node B as the starting node or the ending node.
As for the SA signal, as with the SI1 signal and the SI2 signal, the optical I / F unit of the notification destination may be prompted to shut down the power supply (Shutdown).

FIG. 3 shows details of an optical node device constituting a ring-shaped network system to which the present invention of FIG. 1 is applied. Hereinafter, the description will be given focusing on the difference from the optical node device of FIG.
First, there is a change in the transfer destination port in the label switch unit 20.
The transfer destination port when (1) its own optical node device is the transmitting node and (3) its own optical node device is the destination node is the same as in FIG.
(2) When its own optical node device is a relay node, it is changed so that the receiving port and the transmitting port are the same. For example, the label switch unit 20 transfers the right user packet from the port C distribution unit 23 to the port C confluence unit 24, and transfers the left user packet from the port D distribution unit 21 to the port D confluence unit 22. Transfer to.

Next, in the process of notifying the S-OAM mixing units 32 and 34 of the RDI signal and transmitting the RDI signal flag in the S-OAM packet, the reception error on the first IF11 side is transmitted to the second IF12 side. The reception error on the second IF12 side is notified to the first IF11 side as an RDI signal.
That is, in FIG. 3, the destination where the first port control circuit 11C notifies the RDI signal indicating the failure of the first port 11P is changed from the S-OAM mixing unit 34 to the S-OAM mixing unit 32. Similarly, the destination at which the second port control circuit 12C notifies the RDI signal indicating the failure of the second port 12P is changed from the S-OAM mixing unit 32 to the S-OAM mixing unit 34.
The optical node device that is the source or destination of the RDI signal notifies the administrator of the details of the failure (failure location, failure cause, etc.) indicated by the RDI signal via the administrator's status display terminal or the like. May be encouraged for maintenance work such as component status diagnosis and replacement.

Then, the SI1 signal and SI2 signal in case 1 (IF failure) and the SA signal in case 2 (device failure) are not notified to the optical bypass device outside the optical node device in FIG. 2, but in FIG. Then, the optical bypass mechanism inside the optical node device (bypass switch units 11S, 12S on the optical I / F unit side where the failure has occurred) is notified.
As a result, the optical I / F unit at the time of failure wraps the packet from RX to TX by the folding circuit of the internal optical bypass mechanism, so that the packet is not mixed inside the packet processing unit 10.

FIG. 4 is a hardware configuration diagram of the ring-shaped network system of FIG.
Each optical node device (nodes A, B, C) is a general-purpose device including a CPU (Central Processing Unit) which is a control unit and a storage unit such as a memory, an HDD (Hard Disk Drive), and an SSD (Solid state drive). Configured as a computer.
An accelerator card is connected to the optical node device by a bus standard such as PCI Express. This accelerator card can be equipped with two or more optical I / F units (first IF11, second IF12) such as SFP +. By modularizing the optical I / F section, support for different optical signal standards (for example, different wavelengths and modulation methods) can be realized by exchanging each module, and production according to the common standard reduces the cost of module production. The effect is obtained.

Further, the accelerator card can introduce a config circuit 18 mounted by an FPGA (Field Programmable Gate Array) or the like. The FPGA config circuit 18 has a built-in CRAM (Configuration RAM), and a sequence of CRAM values (binary sequence) describes the attributes of the logic circuit configured in the FPGA and its wiring structure.
The FPGA-based config circuit 18 is just an example of implementation, and a logic circuit may be described in HDL (Hardware Description Language), or a binary string is written to the substrate as a mask ROM (non-rewritable storage element). You may use the one.

FIG. 5 shows a circuit configuration on the accelerator card of FIG. This accelerator card is a virtual configuration of the packet processing unit 10 of the optical node device of FIG. 3 as a config circuit 18 (circuit configuration for FPGA). As a result, it is possible to change the internal wiring only by changing the config setting.
In FIG. 5, in addition to the configuration of FIG. 3, P-OAM removing units 61 and 63, P-OAM mixing units 62 and 64, etherization releasing units 71 and 73, and etherization setting units 72 and 74 are provided. Have. Although the processing unit of the protocol processing unit (PHY layer, MAC layer) for performing Ethernet standard communication is omitted in FIG. 5, the optical I / F unit and the S-OAM processing unit (S-OAM removing unit) are not shown. It is mounted between 31, 33 and S-OAM mixing portions 32, 34).
Further, in FIG. 5, instead of the third IF13 in FIG. 3, the configuration is such that the network interface (PC output, PC input) of the host computer connected by the bus standard is recognized.

The S-OAM packet described above is transmitted and received between two adjacent optical node devices. On the other hand, the P-OAM packet flows through the end-to-end route of (1) transmission node → (2) relay node → (3) destination node, similarly to the user packet. As a result, it is possible to detect where in the data path on the route the abnormality has occurred.
Similar to the user packet, the P-OAM packet is given a label for each user. The P-OAM packet is (1) mixed with the P-OAM mixing units 62 and 64 of the transmitting node and started to be transmitted, (2) the relay node is transferred as it is, and (3) the P-OAM of the destination node is P-OAM. It is removed by the removing units 61 and 63.

The P-OAM removal units 61 and 63 determine a state in which P-OAM packets are not detected for a certain period of time (LOC: Loss of Connection), and automatically select a redundant route using the result.
Here, the P-OAM removing portions 61 and 63 and the P-OAM mixing portions 62 and 64 are located inside the PC (opposite to the optical I / F portion) of the label switch portion 20. Therefore, both the labeled user packet and the P-OAM packet are encapsulated by the etherization setting units 72 and 74 and flow in the network. After that, it is decapsulated by the destination etherization release units 71 and 73.

As a result, the optical node device that detects the interruption of the P-OAM packet can prevent packet loss due to a user packet failure by switching to the "path where the end-to-end path common state is normal". Can be done.
Further, the config circuit 18 is also configured as a detachable component on the accelerator card of FIG. 5, similarly to the optical I / F section.
For example, when the optical I / F unit of FIG. 2 having no optical bypass mechanism is attached, the config circuit 18 also operates as each processing unit such as the packet processing unit 10 of FIG. As a result, the optical node device can be operated even in the conventional configuration in which the optical bypass mechanism is not provided inside the optical node device.
On the other hand, when the optical I / F unit of FIG. 3 having an optical bypass mechanism is attached, the config circuit 18 may also operate as each processing unit such as the packet processing unit 10 of FIG. That is, even with the same optical node device, the old operation (external bypass of the optical node device) and the new operation (internal bypass of the optical node device) can be executed by the detachable parts.

FIG. 6 is a flowchart showing processing related to the P-OAM packet.
First, when the P-OAM removing unit 63 generates the right LOC signal (S101, Yes), the P-OAM mixing unit 64 changes to the RDI mode in which the RDI signal is included in the mixed right P-OAM packet. (S111).
Next, when the P-OAM removing unit 63 generates the left LOC signal (S112, Yes), the P-OAM mixing unit 62 enters the RDI mode in which the RDI signal is included in the mixed left P-OAM packet. Change (S113).
On the other hand, if No in S112, the redundant route selection unit 51 changes to select the left route (S114).

When No in S101, when the P-OAM removing unit 63 generates the left LOC signal (S102, Yes), the P-OAM mixing unit 62 transmits the RDI signal to the mixed left P-OAM packet. Change to the included RDI mode (S103).
Then, the redundant route selection unit 51 is changed to select the right route (S104).

As described above, the internal bypass of the optical node device of FIGS. 3 and 5 made it possible to easily add a redundancy mechanism to the optical node device. Here, the reason why it is difficult to conceive the internal bypass mechanism of the optical node device of FIGS. 3 and 5 is supplemented.

Hereinafter, the features of the optical I / F section of the present embodiment will be described with reference to FIG. 7.
Widely used communication standards such as Ethernet are designed on the assumption that the communication path will pass through the same path in both directions. For example, consider a case where communication including a "call (SYN)-> response (ACK)" such as a TCP (Transmission Control Protocol) handshake is performed between node A and node B.
It is assumed that there are two communication paths (left system and right system) between node A and node B, and one of them is virtually disconnected by STP (Spanning Tree Protocol) or the like. Therefore, it is necessary to communicate with the other side that is not disconnected.

If, as shown in FIG. 7A, a response (ACK) is returned from the same left system to the call (SYN) received from the left system, the response (ACK) cannot be returned due to the disconnection of the left system. Will end up.
On the other hand, as shown in FIG. 7B, if the response (ACK) is returned from the right system in response to the call (SYN) received from the left system, the response (ACK) can be returned even if the left system is disconnected.
That is, the implementation in which one optical I / F unit is dedicated to one system as shown in FIG. 7A has a disadvantage when used alone. On the other hand, in combination with the optical bypass mechanism of the optical node device (of the present embodiment) of FIG. 3, and by explicitly determining the route and setting the path for each system independently by MPLS-TP or the like, redundancy is achieved. Demonstrates the effect of measures against obstacles.

In the present embodiment described above, by using the optical bypass mechanism in combination with the optical node device, the reliability of the entire system does not deteriorate even if the number of nodes in which the optical node devices participating in the network increases.
Furthermore, it is necessary to separately wire and supply power by incorporating the bypass switch units 11S and 12S as the optical bypass mechanism in the optical I / F unit and integrating them with the transmitter (TX) and receiver (RX). Is gone.
Also, since the line bypassed by the optical bypass mechanism is closed inside one optical I / F section (because it does not span multiple optical I / F sections), each optical I / F section is made into an SFP + standard. Can be run along. Therefore, the individual optical I / F units can be freely attached and detached without being affected by the physical layout of the optical node device.

10 Packet processing unit 11 1st IF
11C 1st port control circuit (control unit)
11S Bypass Switch 11SS Optical Bypass Switch 12 2nd IF
12C 2nd port control circuit (control unit)
12S Bypass Switch 12SS Optical Bypass Switch 13 3rd IF
19C controller (control unit)
18 Config circuit 20 Label switch section 21,23,26,28 Distribution section 22,24,25,27 Confluence section 31,33 S-OAM removal section 32,34 S-OAM mixing section 41,43 Label removal section 42,44 Labeling unit 51 Redundant route selection unit 52 Redundant route creation unit 61, 63 P-OAM removal unit 62,64 P-OAM mixing unit 71,73 Etherization release unit 72,74 Etherization setting unit

Claims (6)

An optical node system for an optical node equipment is a plurality mountable optical I / F unit for connecting optical fibers,
The optical node device includes a packet processing unit that receives a first optical I / F unit or a second optical I / F unit as the wearable optical I / F unit.
The optical node system includes an optical bypass device that forms a bypass path outside the optical node device when the second optical I / F unit is mounted on the optical node device.
The first optical I / F section is
A receiver that connects the optical fiber for reception and
A transmitter for connecting the optical fiber for transmission and
It has an optical bypass mechanism configured to straddle a reception signal line input from the optical fiber to the receiver and a transmission signal line output from the transmitter toward the optical fiber.
The second light I / F section is
A receiver that connects the optical fiber for reception and
It has a transmitter for connecting the optical fiber for transmission, and has a transmitter.
The packet processing unit
When the first optical I / F unit is attached, the reception signal line is normally conducted to the receiver and the transmission signal line is conducted to the optical bypass mechanism in the first optical I / F unit. Is conducted to the transmitter, and in the event of an abnormality, a bypass path is formed so as to fold back a signal from the receiving signal line to the transmitting signal line.
When the second optical I / F unit is attached, the light is characterized in that the optical bypass device forms a bypass path so as to bypass the optical fiber for reception to the optical fiber for transmission in the event of an abnormality. Node system. The packet processing unit is configured as a config circuit that can change the internal wiring according to the config setting.
An accelerator card including the config circuit and the attachment / detachment portion of the optical I / F unit is connected to the optical node device.
The bypass path corresponds to different config settings depending on whether the first optical I / F portion is attached to the detachable portion and the second optical I / F portion is attached to the detachable portion. The optical node system according to claim 1, wherein the internal wiring for forming the above is changed. The optical node device further includes a control unit that controls the optical bypass mechanism during normal and abnormal conditions.
Wherein, when the fault in the first optical I / F unit given of the plurality of the first optical I / F section has occurred, the optical bypass of the first optical I / F portion of the predetermined It is characterized in that the mechanism is instructed to form the bypass path, and the other first optical I / F unit is made to transmit a packet for notifying the occurrence of its own failure to another device. The optical node system according to claim 1 or 2. The control unit is attached to its own optical node device when a failure occurs in the packet processing unit inside the optical node device for relaying communication between a plurality of the first optical I / F units. for a plurality of said first optical I / F unit respectively, in claim 3, characterized in that an instruction to form the bypass passage to said first optical I / F portion of the optical bypass mechanism The optical node system described. The set of optical node systems according to any one of claims 1 to 4 has a ring shape that returns from the predetermined optical node system through another optical node system to the predetermined optical node system. The optical fiber network is formed, and the lines between the two optical node systems constituting the set of the optical node systems are redundantly connected by two systems having different signal transmission directions from each other.
Each of the first optical I / F units has the receiver for connecting the optical fiber of one of the two systems and the transmitter for connecting the optical fiber of the one system. A ring-shaped network system characterized by being present. A failure recovery method according to an optical node system for an optical node equipment is a plurality mountable optical I / F unit for connecting optical fibers,
The optical node device includes a packet processing unit that receives a first optical I / F unit or a second optical I / F unit as the wearable optical I / F unit.
The optical node system includes an optical bypass device that forms a bypass path outside the optical node device when the second optical I / F unit is mounted on the optical node device.
The first optical I / F section is
A receiver that connects the optical fiber for reception and
A transmitter for connecting the optical fiber for transmission and
It has an optical bypass mechanism configured to straddle a reception signal line input from the optical fiber to the receiver and a transmission signal line output from the transmitter toward the optical fiber.
The second light I / F section is
A receiver that connects the optical fiber for reception and
It has a transmitter for connecting the optical fiber for transmission, and has a transmitter.
The packet processing unit
When the first optical I / F unit is attached, the reception signal line is normally conducted to the receiver and the transmission signal line is conducted to the optical bypass mechanism in the first optical I / F unit. Is conducted to the transmitter, and in the event of an abnormality, a bypass path is formed so as to fold back a signal from the receiving signal line to the transmitting signal line.
When the second optical I / F unit is attached, an obstacle characterized by forming a bypass path in the optical bypass device so as to bypass the optical fiber for reception to the optical fiber for transmission in the event of an abnormality. How to recover.

Images