JP5351689B2 - Ethernet transfer equipment - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device for restraining wasteful power consumption by adaptively controlling an interface to be used in linkage with a traffic volume to be transferred, in an Ethernet transfer device of 40 GbE or 100 GbE. <P>SOLUTION: The traffic volume of low-speed Ethernet is monitored, and increase and decrease of the number of used lanes of 40 GbE and 100 GbE interfaces is determined in response to the traffic volume. In the case of decrease, the used lanes of the 40 GbE and 100 GbE interfaces are halted, thereafter the used lanes of 40 GbE and 100 GbE interfaces of an opposite device are halted, and thereby the number of lanes is decreased. In the case of increase, unused lanes of the 40 GbE and 100 GbE interfaces are restored, thereafter the unused lanes of the 40 GbE and 100 GbE interfaces of the opposite device are restored, and thereby the number of lanes is increased. <P>COPYRIGHT: (C)2011,JPO&amp;INPIT

Description

  The present invention relates to a technology that contributes to interface speed control and power saving adapted to the amount of traffic transferred by an Ethernet (registered trademark) switch device in the field of communication.

  In the current optical network, the interface is always operated in full operation regardless of the amount of traffic flowing. For this reason, there is a problem that the power of the apparatus is wasted in an interface with a small traffic amount. As an attempt to solve this problem, Non-Patent Document 1 discusses a method for reducing power consumption by changing the speed of the Ethernet interface according to the amount of traffic.

  Standardization of the 40 Gigabit Ethernet (40 GbE) and 100 Gigabit Ethernet (100 GbE) interfaces is underway in the IEEE802.3ba Ethernet Task Force, and the standardization is scheduled to be completed in the fall of 2010. However, the IEEE802.3ba Ethernet Task Force has not studied power saving by adaptive control of the Ethernet interface according to traffic (Non-Patent Document 2).

  Here, the architecture of the 40 GbE / 100 GbE interface currently under consideration for standardization will be described. FIG. 13 shows the architecture of the 40 GbE and 100 GbE interfaces.

  The positioning of this architecture corresponds to the data link layer and the physical layer on the OSI (Open Systems Interconnection) model. A MAC (Media Access Control) sublayer corresponds to the data link layer, and the other portions correspond to the physical layer. Note that FEC (Forward Error Correction) and AN (Auto Negotiation) in FIG. 13 are optional functions and are not essential.

  FIG. 14 is a functional block diagram in PCS (Physical Coding Sublayer). The PCS is a part that converts an Ethernet frame into a bit string at the time of data transfer. The 40GbE / 100GbE PCS has a function of “Block Distribution” and “Alignment Insertion”. These are functions not found in conventional Ethernet.

  FIG. 15 shows the function of Block Distribution. As shown in FIG. 15, a 64B / 66B encoded signal in PCS is distributed to each PCS lane by a round robin method. The number of PCS lanes differs between 40 GbE and 100 GbE. The number of PCS lanes for 40 GbE is 4, and the number of PCS lanes for 100 GbE is 20.

  FIG. 16 shows the function of Alignment Insertion. FIG. 16a shows the insertion of the alignment insertion, and FIG. 16b shows the interval of the alignment insertion. As shown in FIG. 16, an alignment marker is inserted for every 16383 64B / 66B encoded signals.

  FIG. 17 shows a configuration of 40 GbE and 100 GbE PMA (Physical Media Attachment). In the case of 40 GbE, the data distributed to the four lanes in the PCS is transferred to the physical interface via the PMA without being multiplexed. On the other hand, in the case of 100 GbE, the signals distributed to 20 lanes in the PCS are multiplexed from 20 to 10 and transferred to the physical interface. In some cases, 10 signals are further multiplexed into 4 signals and transferred to the physical interface.

IEEE802.1az, “Energy Efficient Ethernet” Draft 2.0 IEEE802.3ba, “40Gb / s and 100Gb / s Ethernet Task Force” Draft 2.1

  However, the interface to which the method studied in Non-Patent Document 1 is applied is (1) 100BASE-T (Full duplex), (2) 1000BASE-T (Full duplex), (3) 1000BASE-KX, (4) Limited to 10GBASE-T / KR / KX4 (Remarks: KR and KX indicate equipment backplane), and application to 40 Gigabit Ethernet and 100 Gigabit Ethernet is not considered. In addition, the speed changing method of parallel Ethernet signals (for example, 10GBASE-KX4) targeted by Energy-Efficient Ethernet (EEE) of Non-Patent Document 1 cannot be applied as it is to 100 GbE of multilane signals.

  Therefore, an object of the present invention is to provide a device that suppresses wasteful power consumption by adaptively controlling an interface to be used in association with a traffic amount to be transferred in a 40 GbE or 100 GbE Ethernet transfer device.

The Ethernet transfer means according to the present invention for realizing the above object is an Ethernet transfer apparatus having a low-speed Ethernet interface for transmitting / receiving low-speed Ethernet and a multi-lane high-speed Ethernet interface for transmitting / receiving high-speed Ethernet. Monitoring means; means for determining an increase / decrease in the number of used lanes of the high-speed Ethernet interface according to the traffic volume; a used lane number reducing means; and a used lane number increasing means ; When reducing the number of used lanes, change the number of lanes used on the transmission side of the high-speed Ethernet interface, then pause the unused lanes on the receiving side of the high-speed Ethernet interface of the opposite device, and change the number of lanes used on the transmission side. Thereafter, the transmission-side unused lane and the reception-side unused lane of the high-speed Ethernet interface are paused, and then the transmission-side unused lane of the high-speed Ethernet interface of the opposite device is paused, and the number of used lanes is increased. Means for recovering the transmission-side pause lane and the reception-side pause lane of the high-speed Ethernet interface, and then restoring the transmission-side pause lane and the reception-side pause lane of the high-speed Ethernet interface of the opposite device when the number of used lanes increases. It is characterized by being.

Moreover, it is also preferable that the use lane number reducing unit and the use lane number increasing unit exchange a message including information on the use lane number with the opposite device.

  The message is preferably an operation code of an Ethernet OAM frame.

The use lane number reduction means and the use lane number increase means preferably pause or restore the PCS lane, PMA part, PMD part, and MDI part of the high-speed Ethernet interface.

  The high-speed Ethernet interface is preferably a 40 Gigabit Ethernet interface or a 100 Gigabit Ethernet interface.

  According to the present invention, it is possible to provide an Ethernet transfer apparatus that suppresses wasteful power consumed by the apparatus by adaptively controlling the speed of the interface to be used according to the traffic volume of the Ethernet signal.

The usage example of the Ethernet transfer apparatus of this invention is shown. The internal structure of an Ethernet transfer apparatus is shown. The structure of a 40 GbE interface is shown. The structure of a 100 GbE interface when the number of signals after multiplexing is 10 is shown. The structure of a 100 GbE interface when the number of signals after multiplexing is four is shown. The sequence at the time of the lane number reduction between opposing Ethernet transfer apparatuses is shown. The sequence when the number of lanes between opposing Ethernet transfer devices increases is shown. FIG. 8 is a flowchart detailing the operation on the initiator side in the sequences of FIGS. 6 and 7. FIG. FIG. 8 is a flowchart detailing the operation on the slave side of the sequences of FIGS. 6 and 7. FIG. The flow of data at the time of data transfer in the case of using one wavelength in a 40 GbE 4-wavelength device is shown. The flow of data during data transfer in the case of using two wavelengths in a 100 GbE four-wavelength device is shown. A data flow at the time of data transfer in the case of using one wavelength in a 100 GbE 10-wavelength device is shown. The architecture of the 40 GbE and 100 GbE interfaces is shown. The functional block diagram in PCS (Physical Coding Sublayer) is shown. Indicates the function of Block Distribution. Indicates the function of Alignment Insertion. The structure of 40 GbE and 100 GbE PMA (Physical Media Attachment) is shown.

  The best mode for carrying out the present invention will be described in detail below with reference to the drawings. FIG. 1 shows an example of use of the Ethernet transfer apparatus of the present invention. The link between the Ethernet transfer apparatuses 1 is connected point-to-point, and 40 GbE or 100 GbE is used as the link between them. Assume that the Ethernet transfer device 1 aggregates 1 GbE, 10 GbE, and the like and transfers them at 40 GbE or 100 GbE.

  FIG. 2 shows an internal configuration of the Ethernet transfer apparatus. The difference between FIG. 2a and FIG. 2b is the interface part. FIG. 2 a shows a case with a 40 GbE interface 14, and FIG. 2 b shows a case with a 100 GbE interface 16.

  The low-speed Ethernet interface 11 has a function of transmitting and receiving a lower-speed Ethernet signal than a high-speed interface such as 1 GbE or 10 GbE.

  The Ethernet signal aggregating unit 12 has a function of aggregating signals such as 1 GbE and 10 GbE received via the Ethernet interface.

  The 40 GbE signal generation unit 13 and the 100 GbE signal generation unit 15 have a function of generating the Ethernet signals aggregated by the Ethernet signal aggregation unit 12 as 40 GbE signals and 100 GbE signals. In addition, this signal generation part has a function which distributes a signal to Physical Coding Sublayer (PCS).

  The 40 GbE interface 14 and the 100 GbE interface 16 have a function of converting a 40 GbE signal distributed to the PCS or a 100 GbE signal into each medium and transmitting / receiving it. Here, the 40 GbE interface 14 and the 100 GbE interface 16 have multilane interfaces.

  FIG. 3 shows the configuration of the 40 GbE interface. In the case of 40 GbE, it has four PCS lane numbers 0 to 3, and the number of lanes of signals transferred from the interface after media conversion (electric-optical conversion) is also four. On the receiving side, each of the four signals is received and sent to the PCS lane after media conversion.

  FIG. 4 shows the configuration of a 100 GbE interface when the number of multiplexed signals is 10. In this case, the number of PCS lanes is 20, and the number of lanes of signals transferred from the interface after the PCS signals are multiplexed in PMA (Physical Media Attachment) is 10. Here, two PCS lane signals are alternately multiplexed in the PMA two by ten to form ten signals. These 10 signals are transmitted to the opposite device after being converted into optical or electrical signals according to each transfer medium. On the receiving side, processing opposite to that on the transmitting side is performed. That is, 10 signals are received, and each signal is separated into two signals and then sent to the PCS lane.

  FIG. 5 shows the configuration of a 100 GbE interface when the number of multiplexed signals is four. In this case, the number of lanes of the signal transferred from the interface after the PCS signal is multiplexed in PMA (Physical Media Attachment) is four. Here, first, two PCS lane signals are alternately multiplexed in the PMA to become 10 signals, and then these 10 signals are further multiplexed in order in the PMA to obtain four signals. It becomes. The four signals are converted into optical or electrical signals according to each transfer medium, and then transmitted to the opposite device. On the receiving side, processing opposite to that on the transmitting side is performed. That is, after receiving 4 signals and separating them into 10 signals, each signal is separated into 2 signals and sent to the PCS lane.

  Here, an Ethernet OAM frame used for notification between Ethernet transfer apparatuses will be described (reference document: Japanese Patent Laid-Open No. 2008-131614 and ITU-T T.1731 recommendation).

As a layer 2 network for a LAN (Local Area Network), Ethernet technology is widely used. A network service provider needs to perform operation administration and maintenance (OAM) from a remote location in order to always provide a stable service to subscribers. For this purpose, Ethernet OAM frames are defined by ITU-T Y.1731 or IEEE802.1ag “Connectivity Fault Management”. In particular, ITU-T Y.1731 “OAM functions and Mechanisms for Ethernet based Networks” is a standard that realizes operation and maintenance management functions almost equivalent to SONET / SDH (Synchronous Optical NETwork / Synchronous Digital Hierarchy) in Ethernet networks. A general Ethernet OAM frame configuration is shown below.

  The Ethernet OAM frame is a frame having functions such as reachability management (CFM: Connectivity Fault Management), error notification, link (line) performance monitor, and the like and a function for transmitting the information. The Ethernet OAM frame includes a destination MAC (Media Access Control) address, a source MAC address, an Ethernet OAM TLV (Type Length Value), and an FCS (Frame Check Sequence). The MAC address is a 48-bit identification number that is unique to the layer 2 network device that is the target of operation and maintenance management or unique to the network interface card. The Ethernet OAM TLV includes, for example, control information (request / response), status (power status, received light status, link disconnection, failure, etc.), vendor code, model code, and the like.

  FIG. 6 shows a sequence when reducing the number of lanes between the opposing Ethernet transfer apparatuses, and FIG. 7 shows a sequence when increasing the number of lanes between the opposing Ethernet transfer apparatuses. FIG. 8 shows a detailed flowchart of the operation on the initiator side (device for starting lane number control) in the sequences of FIGS. 6 and 7, and FIG. 9 shows the slave side (initiator side) of the sequences in FIGS. The flowchart which detailed operation | movement of the apparatus which performs lane number control in response to this operation | movement is shown.

The operation on the initiator side is shown below according to FIG.
S1. Traffic volume monitoring: The traffic volume on the LAN side is monitored, and the increase or decrease in the number of lanes of the 40 GbE / 100 GbE signal is determined according to the traffic volume. If the number of lanes is to be reduced, the process proceeds to Step 2, and if it is increased, the process proceeds to Step 9. Note that the traffic volume monitoring and the link number increase / decrease determination method are calculated based on MIB (Management Information Base) information on the number of packets transferred between L2SWs (reference document: Japanese Patent Application Laid-Open No. 2006-157102). ). Here, the ratio of inflow traffic is expressed as follows.
Inflow traffic [%] = (number of inflow bytes × 8) × 100 / (interface speed × monitoring time interval [s])

Here, the interface speed is 40 Gb / s or 100 Gb / s. Table 2 shows an example of determining the number of lanes according to the traffic volume. For example, when the traffic is 20% or more and less than 30%, the number of 40 GbE lanes is 2, the number of 100 GbE lanes is 3 (in the case of 10 lanes), or 2 (in the case of 4 lanes).

  As for the number of lanes to be allocated according to traffic, a method of taking the number of lanes more than the above table is also effective in order to provide a sufficient capacity for the traffic. This is determined according to the operation policy.

"When reducing the number of lanes"
S2. Transmission side use lane change: Data is distributed to the use lane determined in step 1. Here, the number of lanes through which data flows is controlled according to the traffic volume monitored in step 1. For example, in 100 GbE (in the case of 10 lanes) in Table 2, when the traffic volume becomes 45%, the number of lanes becomes 5. At this time, the 64B / 66B encoded signal is distributed to PCS lanes 0 to 9 by the round robin method. Next, data is transferred using five lanes transmitted by multiplexing the signals of PCS lanes 0 to 9. In the remaining lanes, empty data is transferred.

  Here, for 40 GbE (4 lanes) and 100 GbE (4 lanes or 10 lanes), the PCS lanes used for each lane number are shown below.

For example, in Table 3 “40 GbE (4 lanes)”, when the use lane (wavelength) is (1, 2), the PCS lane to be used is (0, 1). Similarly, in Table 4 “100 GbE (10 lanes)”, for example, when the use lane (wavelength) is (1, 2, 3, 4, 5), the PCS lane to be used is (0, 1, 2). , 3, 4, 5, 6, 7, 8, 9). In Table 5 “100 GbE (4 lanes)”, the PCS lane to be used is irregular as shown in the table in accordance with each used lane (wavelength).

S3. Notification of number of used lanes: The number of used lanes is notified to the opposite device. For this notification, the above-described Ethernet OAM frame OpCode is used.
S4. Wait for notification: Waits for a notification of setting completion from the opposite device. If the notification is confirmed, the process proceeds to the next process. If the notification cannot be confirmed within a certain time, it is considered that a failure has occurred, a failure notification is sent to the host device, etc., and this control is stopped.
S5. Non-use of transmission side lane: A transmission side lane that is no longer used is paused by changing the number of used lanes in step 2. In the case of lane suspension, the PCS lane, PMA portion, PMD portion, and MDI portion are suspended.
S6. Receiving side unused lane pause: Pauses a receiving side lane that is no longer used. Again, the PCS lane, PMA part, PMD part, and MDI part are paused.
S7. Setting completion notification: Notifies the opposite device that the setting has been completed.
S8. Wait for notification: Waits for a notification of setting completion from the opposite device. If the notification is confirmed, the lane number reduction process is terminated. If the notification cannot be confirmed within a certain time, it is considered that a failure has occurred, a failure notification is sent to the host device, etc., and this control is stopped.

"In the case of an increase in the number of lanes"
S9. Restoration lane restoration on transmission side: A lane to be restored is determined according to the traffic amount monitored in step 1, and the lane is restored. At this time, empty data is transmitted. In the case of lane restoration, the PCS lane, PMA part, PMD part, and MDI part are restored.
S10. Receiving side sleep lane recovery: A lane to be recovered is determined according to the traffic volume monitored in step 1, and the lane is recovered. In this case, the PCS lane, PMA part, PMD part, and MDI part are restored.
S11. Notification of number of used lanes: The number of used lanes is notified to the opposite device. For this notification, the above-described Ethernet OAM frame OpCode is used.
S12. Wait for notification: Waits for a notification of setting completion from the opposite device. If the notification is confirmed, the process proceeds to the next process. If the notification cannot be confirmed within a certain time, it is considered that a failure has occurred, a failure notification is sent to the host device, etc., and this control is stopped.
S13. Reception side recovery lane reception start: Reception using the lane recovered in step 10 is started.
S14. Transmission side restoration lane transmission start: For the lane restored in step 9, the 64B / 66B encoded signal is distributed by the round robin method.
S15. Setting completion notification: Notifies the opposite device that the setting has been completed.
S16. Wait for notification: Waits for a notification of setting completion from the opposite device. If the notification is confirmed, the lane number increase process is terminated. If the notification cannot be confirmed within a certain time, it is considered that a failure has occurred, a failure notification is sent to the host device, etc., and this control is stopped.

The operation on the slave side is shown below according to FIG.
S21. Operation code monitoring: 40GbE or 100GbE interface traffic is monitored and the OpCode of the Ethernet OAM frame is confirmed among the received packets. The Ethernet OAM frame can be identified by checking the value of the Length / Type field of the packet and the value of the Subtype field (the values of each field are ITU-T Y.1731 “OAM functions and Mechanisms for Ethernet based Networks ”). In accordance with the operation specified by the received OpCode, the control shifts to the number of lanes to be used. If the number of lanes is to be reduced, the process proceeds to step 22, and if it is increased, the process proceeds to step 29.

"When reducing the number of lanes"
S22. Downlink traffic volume reduction confirmation: Checks whether the downlink traffic volume is decreasing in the same way as the upstream traffic volume. Since Ethernet is vertically asymmetric, even if the amount of upstream traffic decreases, the amount of downstream traffic does not necessarily decrease. If the number of lanes is reduced while the amount of downlink traffic is not decreasing, network congestion or packet loss may occur due to a decrease in bandwidth. Therefore, this control is stopped by transmitting a reset signal to the opposite device without reducing the number of lanes. Note that the opposite device (initiator side) that has received the reset signal also stops reducing the number of lanes. If the downstream traffic volume has decreased to be equal to or higher than the upstream traffic volume, the process proceeds to the next process.
S23. Receiving side unused lane pause: In accordance with the operation code confirmed in step 21, the receiving side unused lane is paused. In the case of lane suspension, the PCS lane, PMA portion, PMD portion, and MDI portion are suspended.
S24. Transmission side use lane change: According to the operation code confirmed in step 21, the number of use lanes on the transmission side is changed. The lane number changing process is performed in the same manner as the process in step 2 on the initiator side.
S25. Setting completion notification: Notifies the opposite device that the setting has been completed.
S26. Wait for notification: Waits for a notification of setting completion from the opposite device. If the notification is confirmed, the process proceeds to the next process. If the notification cannot be confirmed within a certain time, it is considered that a failure has occurred, a failure notification is sent to the host device, etc., and this control is stopped.
S27. Pause unused lane pause: Pauses the lane interface that is transferring empty data. Again, the PCS lane, PMA part, PMD part, and MDI part are paused.
S28. Setting completion notification: The opposite device is notified that the setting has been completed, and the lane number reduction processing ends.

"In the case of an increase in the number of lanes"
S29. Receiving-side pause lane recovery: The receiving-side pause lane is restored according to the operation code of step 21. In the case of lane restoration, the PCS lane, PMA part, PMD part, and MDI part are restored.
S30. Restoration lane recovery on transmission side: In response to the operation code in step 21, the pause lane on the transmission side is restored. At this time, empty data is transmitted. In this case, the PCS lane, PMA part, PMD part, and MDI part are restored.
S31. Setting completion notification: Notifies the opposite device that the setting has been completed.
S32. Wait for notification: Waits for a notification of setting completion from the opposite device. If the notification is confirmed, the process proceeds to the next process. If the notification cannot be confirmed within a certain time, it is considered that a failure has occurred, a failure notification is sent to the host device, etc., and this control is stopped.
S33. Reception side recovery lane reception start: Data reception using the reception side recovered lane is started.
S34. Setting completion notification: The opposite device is notified that the setting has been completed, and the lane number reduction processing ends.

  The status notification means between apparatuses will be described below. In the present invention, an operation code and a code indicating the number of used lanes are included in an Ethernet OAM (Operation Administration and Maintenance) frame. As the code, a code reserved for future use in the OpCode of the Ethernet OAM frame is used. Currently 2,4,7-31,64-255 are reserved for IEEE. 33-63 is reserved for ITU use. The Ethernet OAM frame is used to notify the status between devices.

Examples of operation code types are shown below. The following is an example, and any operation code may be used as long as it is an empty code.
Operation code: Operation: Number of wavelengths used
33: Change in number of used lanes Number of used wavelengths Use wavelength 1 out of 4 wavelengths of 40GbE 34: Change in number of used lanes Use of wavelength 1-2 out of 4 wavelengths of 40GbE 35: Change in number of used lanes Number of used wavelengths 40GbE 36: Use lane number change Use wavelength number Use 40GbE wavelength 4 Use wavelength 1-4: Use lane change Use wavelength number Use 100GbE wavelength 1 out of 4 wavelengths Use 38: Change in the number of used lanes Use the number of used wavelengths 1-2 of the 100GbE wavelengths used 39: Change the number of used lanes Use the number of used wavelengths 100-3 of the 100GbE wavelengths 1-3 Use 40: Change the number of used lanes Use Number of wavelengths Use wavelength 1-4 out of 4 wavelengths of 100 GbE 41: Change number of used lanes Use number of wavelengths Use wavelength 1 out of 10 wavelengths of 100 GbE 42: Number of used lanes changed Number of used wavelengths Use wavelength 1-2 out of 10 wavelengths of 100 GbE 43: Changed number of used lanes Used number of wavelengths Used out of 10 wavelengths of 100 GbE Use wavelength 1-3 44: Changed number of used lanes Used Number of used wavelengths 100 GbE 45: Use lane number is changed 45: Use lane number is changed Use wavelength number 1 is used among 100 GbE 10 wavelengths 46: Use lane number is used Use wavelength number is used Wavelength 1 is out of 10 wavelengths of 100 GbE 6 is used 47: Number of used lanes is changed. Number of used wavelengths is used. Wavelength 1-7 is used out of 10 wavelengths of 100 GbE. 48: Number of used lanes is changed. Number of used wavelengths is wavelength 1-8 is used among 10 wavelengths of 100 GbE. 49: Number of used lanes. Change Number of used wavelengths Use wavelength 1-9 out of 10 wavelengths of 100 GbE 50: Change number of used lanes Number of used wavelengths 10 waves of 100 GbE Using the wavelength 1-10 of 51: setting completion notification 52: Reset

The switch operation at the time of status notification will be described below.
(I) In the case of a 40GbE 4-wavelength device When using one of the four wavelengths (in the case of the operation code 33)
In this case, the following operation is performed.
Transmitting side: Data is transferred to 0 in the PCS lane.
Receiver side: Transfers data to 0 of PCS lane.
FIG. 10 shows the data flow during data transfer. The dark gray part indicates the part used for data transfer. The light gray part indicates a resting point. (II) In the case of a 100 GbE 4-wavelength device When two wavelengths (wavelengths 1 & 2) out of four wavelengths are used (in the case of the operation code 38)
In this case, the following operation is performed.
Transmission side: Data is transferred to 0, 1, 4, 5, 8, 9, 12, 13, 16, 17 of the PCS lane.
Reception side: Data is transferred to 0, 1, 4, 5, 8, 9, 12, 13, 16, and 17 of the PCS lane.
FIG. 11 shows the data flow during data transfer.
(III) In the case of a 10 GbE 10-wavelength device When one wavelength (wavelength 1) out of 10 wavelengths is used (in the case of the operation code 41)
In this case, the following operation is performed.
Transmission side: Data is transferred to 0 and 1 of the PCS lane.
Receiver: Transfers data to 0 and 1 of PCS lane.
FIG. 12 shows the data flow during data transfer.

  By controlling as described above, wasteful power consumption can be suppressed by controlling the interface used by the Ethernet transfer apparatus in conjunction with the traffic volume.

  Moreover, all the embodiments described above are illustrative of the present invention and are not intended to limit the present invention, and the present invention can be implemented in other various modifications and changes. Therefore, the scope of the present invention is defined only by the claims and their equivalents.

DESCRIPTION OF SYMBOLS 1 Ethernet transfer apparatus 11 Low-speed Ethernet interface 12 Ethernet signal aggregation part 13 40GbE signal generation part 14 40GbE interface 15 100GbE signal generation part 16 100GbE interface

Claims (5)

In an Ethernet transfer device having a low-speed Ethernet interface for transmitting / receiving low-speed Ethernet and a multi-lane high-speed Ethernet interface for transmitting / receiving high-speed Ethernet,
Means for monitoring the traffic amount of the low-speed Ethernet;
Means for determining an increase or decrease in the number of lanes used in the high-speed Ethernet interface according to the traffic volume;
Means to reduce the number of lanes used ;
A means for increasing the number of lanes used ,
The use lane number reducing means is:
In the case of reducing the number of used lanes, change the number of lanes used on the transmission side of the high-speed Ethernet interface, then pause the unused lane on the receiving side of the high-speed Ethernet interface of the opposite device, and change the number of lanes used on the transmission side. The high-speed Ethernet interface transmitting side unused lane and the receiving side unused lane are paused, and thereafter the high-speed Ethernet interface transmitting side unused lane of the opposite device is paused,
The means for increasing the number of lanes used is
In the case of an increase in the number of lanes used, the transmission side pause lane and the reception side pause lane of the high-speed Ethernet interface are restored, and then the transmission side pause lane and the reception side pause lane of the high-speed Ethernet interface of the opposite device are restored. Ethernet transfer device characterized by this. 2. The Ethernet transfer apparatus according to claim 1, wherein the use lane number reducing unit and the use lane number increasing unit exchange a message including information on the use lane number with the opposite device. 3. The Ethernet transfer apparatus according to claim 2 , wherein the message is an operation code of an Ethernet OAM frame. The use lane number reducing means and the use lane number increasing means, PCS lane, PMA part of the Fast Ethernet interface, PMD section, and any one of claims 1, characterized in that pause or recover the MDI portion 3 Ethernet transfer device described in 1. The Fast Ethernet interface, Ethernet transport device according to any one of claims 1 to 4, characterized in that a 40 Gigabit Ethernet interface, or 100 gigabit Ethernet interface.

Images