CN115550281A - Resource scheduling method and framework for AWGR optical switching data center network - Google Patents
Resource scheduling method and framework for AWGR optical switching data center network Download PDFInfo
- Publication number
- CN115550281A CN115550281A CN202211520300.3A CN202211520300A CN115550281A CN 115550281 A CN115550281 A CN 115550281A CN 202211520300 A CN202211520300 A CN 202211520300A CN 115550281 A CN115550281 A CN 115550281A
- Authority
- CN
- China
- Prior art keywords
- rack
- data
- module
- switch
- awgr
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
- H04L47/62—Queue scheduling characterised by scheduling criteria
- H04L47/625—Queue scheduling characterised by scheduling criteria for service slots or service orders
- H04L47/6275—Queue scheduling characterised by scheduling criteria for service slots or service orders based on priority
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
- H04L47/62—Queue scheduling characterised by scheduling criteria
- H04L47/625—Queue scheduling characterised by scheduling criteria for service slots or service orders
- H04L47/6255—Queue scheduling characterised by scheduling criteria for service slots or service orders queue load conditions, e.g. longest queue first
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
- H04L47/62—Queue scheduling characterised by scheduling criteria
- H04L47/6295—Queue scheduling characterised by scheduling criteria using multiple queues, one for each individual QoS, connection, flow or priority
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
- H04Q11/0062—Network aspects
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
- H04Q11/0062—Network aspects
- H04Q2011/0075—Wavelength grouping or hierarchical aspects
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04Q—SELECTING
- H04Q11/00—Selecting arrangements for multiplex systems
- H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
- H04Q11/0062—Network aspects
- H04Q2011/0086—Network resource allocation, dimensioning or optimisation
Abstract
The invention discloses a resource scheduling method facing an AWGR optical switching data center network, which comprises the steps of collecting flow information through a scheduler, sequencing the flow information, analyzing and judging the flow information according to the sequencing, determining whether the current flow can be sent or not, adding the flow information of the part into a result list if the current flow can be sent, finally confirming the length of the result list, and sending a scheduling instruction after the length is determined; the corresponding framework comprises a rack, a dispatcher connected with the rack and an array waveguide grating, wherein the rack comprises a top rack switch and a plurality of servers, and the servers realize communication through a data exchange topology of the top rack switch. The scheduling method solves the problems of optical data packet conflict and limited optical wavelength resources in an optical switching network by allocating time slots and wavelengths, solves the problems of timely allocation of optical wavelength resources and retransmission delay of optical data conflict packets by scheduling strategies, and avoids the waiting delay brought by a request-response process by adopting a staggered time slot method.
Description
Technical Field
The invention relates to the technical field of communication, in particular to a resource scheduling method and a resource scheduling architecture for an AWGR optical switching data center network.
Background
With the rapid development of cloud computing, mobile internet and streaming media, the network traffic of a data center is increasing exponentially, which puts higher and higher requirements on the performance of the data center. The current data center is formed by connecting tens of thousands or even millions of servers together by an electric switch, and the servers meet the real-time dynamic demands of users through frequent interaction and data storage. Due to the limitation of the packaging technology, the bandwidth acceleration of the I/O port of the electric switch is far lower than the rapidly-increased network flow requirement; meanwhile, the traditional data center adopts a network architecture of interconnection of electric switches, and frequent optical-electrical-optical conversion is needed in the communication process between the switches, which brings huge energy consumption. When passing through each stage of switch, the data packet needs to undergo multi-stage queuing and processing latency, which greatly increases data transmission latency. Therefore, the data center network of the conventional electrical switching architecture faces a huge challenge brought by rapidly increasing traffic in terms of switching nodes and network structures, and in order to overcome the above disadvantages, a new optical switching data center network architecture is gradually popular.
Due to the transparent characteristic of a data modulation format and a transmission rate, the optical switch has a large I/O port switching bandwidth, and compared with a traditional multi-layer system architecture of electrical switching, the flat optical switching architecture with highly distributed control shortens the processing time of a controller while providing large capacity and expandability, so that the overall performance of a network is optimized in the aspects of throughput, delay and the like. An Arrayed Waveguide Grating (AWGR) is a passive optical device that is cyclically routed from a given input port wavelength through an output port, and used in conjunction with a fixed wavelength laser to form a data center optical interconnect switching fabric. Because of the missing of the optical cache, when one destination port has two or more source ports (wavelengths) to send data simultaneously in the transmission process of the data, the optical switching data center network based on the AWGR will have the phenomena of optical collision and packet loss at the destination port; in a small-scale data center optical network, a fully-connected form can be adopted, that is, each transceiving end sets an exclusive wavelength for data transmission, but because of the characteristic limitation of an Arrayed Waveguide Grating (AWGR), the number of ports of the Arrayed Waveguide Grating (AWGR) is limited, and a fully-connected network structure cannot interconnect a large-scale data center network.
To solve the above problem of wavelength collision, a scheduling scheme is needed to schedule data transmission in the AWGR to avoid the problem of data packet collision, and to meet the needs of data center scale growth and avoid the limitation of limited wavelength resources, a time slot division manner is used to guide the ordered transmission of wavelengths. Some current static scheduling schemes use a polling method to schedule the time period and wavelength for forwarding, that is, assuming that n wavelengths can transmit data and m ports need to transmit data (n < m), then the 1 st to n th ports of the m ports are selected to transmit data in this time period, then the n +1 th to n + n ports are selected to transmit data in the next time period, and when the last transmission port is selected, the cycle starts from the first transmission port.
The full connection mode has the following defects: the wavelength conflict problem of the AWGR data center network can be solved by using a full connection mode, that is, each transceiver end has an exclusive physical wavelength channel for interconnection, but this interconnection mode would occupy a large number of ports of the AWGR. Due to the limitation of the manufacturing process, the number of ports of the Arrayed Waveguide Grating (AWGR) cannot be increased without limit along with the increase of the network scale, and the interconnection mode can only be applied to a micro data center network.
The static scheduling scheme has the following defects: in a static scheduling scheme, a polling mode is used for arranging and forwarding time slots and wavelengths, each port which is likely to conflict can be allocated with different time slots or wavelengths as much as possible, the scheduling mode cannot allocate optical wavelength resources to the most-needed transmission data in the highest cache with the occupation ratio in time, and cannot meet the real-time requirements of network burst traffic of a data center network on network bandwidth (time slots and wavelengths) resources. Meanwhile, in the request-response process in the scheduling scheme, data transmission needs to wait for the response process, which increases the waiting time delay during data transmission.
Chinese patent CN105959163A discloses a passive optical interconnection network structure and a data communication method based on software definition, which can reduce power consumption, delay and reliability, but the application scenario is limited and the real-time requirement cannot be guaranteed.
Disclosure of Invention
In view of the above problems, an object of the present invention is to provide a resource scheduling method and architecture for an AWGR-oriented optical switching data center network, which are used to solve the problem of wavelength conflict in the AWGR-based optical switching data center, the problem that the optical bandwidth cannot be flexibly allocated on demand in a static scheduling scheme, and the waiting delay caused by the "request-response" process in the scheduling scheme. In order to achieve the purpose, the invention provides the following technical scheme: a resource scheduling method and architecture facing to an AWGR optical switching data center network comprises the following steps:
s1, a top rack switch collects the occupation condition of each cache of each group and a source node and a destination node corresponding to the cache, then sends a request to a scheduler and waits for receiving a replied command, and the scheduler receives the request sent by each top rack switch and analyzes flow information;
s2, the scheduler stores the counted flow information through a priority queue, the priority queue sorts according to the flow of each flow, and the flow is arranged in front of the priority queue;
s3, traversing the elements of the priority queue in sequence, analyzing and judging the stream, determining whether the current stream can be sent, if so, performing the next step, otherwise, re-performing the S3;
s4, adding the current flow into the result array, stopping calculation if the length of the result array reaches the upper limit of the number or the priority queue is empty, and otherwise, re-executing S3;
and S5, the scheduler issues control commands to each top rack switch according to the information of the result array, and the top rack switch controls corresponding cache in the sending module to send data in a staggered time slot mode after receiving the corresponding commands of the scheduler.
Preferably, different buffers in the overhead switch are grouped, and the buffers share one sending module for sending, so that streams sent to different destination terminals are sent in different time slots.
Preferably, the staggered time slot specifically includes: and starting to transmit data at the time t1 of the time slot N, finishing data transmission at the time t2, collecting the cache by a cache module of the top rack switch at the time and transmitting a request to a scheduler, finishing scheduling by the scheduler at the time t3 and transmitting a transmission command to each top rack switch, and transmitting data of the time slot N +1 at the time t3 so as to circulate until finishing.
Preferably, the step S3 further includes traversing the elements of the priority queue according to the order, through two dictionary array maps T And map R Saving the states of the sending module and the receiving module in each top-of-rack switch if map T And map R The value of the element in the system is 0, which indicates that the sending module or the receiving module is not occupied at present and can send; if map T Or map R The value of the element in the system is 1, which indicates that the sending module or the receiving module is occupied, and then the sending module and the receiving module cannot send the data, and the quantity of the sending module and the receiving module is the same.
Preferably, S4 specifically is: if the length of the result array reaches MxN, the data stream which can be sent reaches the upper limit of the number, or the length of the array of the stream is not enough for MxN at the moment, but the priority queue is empty, the calculation is stopped; the number of the top rack switches is N, the number of the sending modules in the top rack switches is M, and the number of streams which can be sent simultaneously is at most M multiplied by N.
The architecture of the resource scheduling method facing the AWGR optical switching data center network comprises a rack, a scheduler and an array waveguide grating, wherein the scheduler and the array waveguide grating are connected with the rack, the rack comprises a top-of-rack switch and a plurality of servers, and the servers are communicated through a data switching topology of the top-of-rack switch.
Preferably, the arrayed waveguide grating structure is a 4 × 4 array, the arrayed waveguide grating routes optical signals to corresponding output ports in a cyclic wavelength routing manner, 4 input ports arranged inside the arrayed waveguide grating input signals with different wavelengths, the number of the transmission modules of each overhead switch is fixed, each transmission module in the same overhead switch can only transmit a unique wavelength λ different from other transmission modules, the arrayed waveguide grating is used as an optical switching node, so that the signals with different wavelengths of each input port can be transmitted to the fixed output ports, and the transmission module is responsible for transmitting data packets in the cache module in the form of optical signals with different wavelengths;
wherein:
λ in (b) indicates that the signal is an optical signal having a specific wavelength, and the upper corner i is the input port number of the signal, and the lower corner w is the wavelength number of the signal.
Preferably, the internal structure of the top-rack switch includes an ethernet switching module, a cache module is disposed between the ethernet switching module and the sending module and connected to the receiving module, a cache module is disposed between the ethernet switching module and the server and connected to the server, data of the server is uploaded to the ethernet switching module from the server in the rack, the ethernet switching module distributes the data according to a destination of the data packet, if the destination is the server in the rack, the data is directly forwarded to the destination server in the rack, and if the destination is the server outside the rack, the data is uploaded to the cache module corresponding to the sending module, and a control instruction is waited for data sending.
Preferably, the number of the sending modules = number of top-of-rack switches/number of time slots = number of wavelengths;
the number of the cache modules = the number of top switches/number of wavelengths = the number of arrayed waveguide gratings;
the number of wavelengths = the number of arrayed waveguide grating ports.
Compared with the prior art, the invention has the beneficial effects that:
for the allocation of time slots and wavelengths, the problems of optical data packet collision and limited optical wavelength resources in an optical switching network are solved; the invention solves the problem of conflict packet retransmission time delay caused by timely distribution of optical wavelength resources and contention of optical data packets, which can not be solved in a static scheduling scheme, by a unique scheduling strategy, and avoids the waiting time delay caused by a request-response process in the traditional scheduling scheme by adopting a staggered time slot method.
Drawings
Fig. 1 is a flowchart of a method for scheduling resources of a central network according to the present invention.
Fig. 2 is a view showing the internal structure of the housing of the present invention.
Fig. 3 is an architecture diagram of the AWGR-oriented optical switching data center network of the present invention.
Fig. 4 is a schematic optical signal flow diagram of a 4x4 arrayed waveguide grating of the present invention.
Fig. 5 is an internal structure diagram of the ethernet switching module according to the present invention.
Fig. 6 is a diagram comparing the misaligned time slots of the present invention and the conventional scheduling time slots.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicate orientations or positional relationships based on the orientations or positional relationships shown in the drawings, and are only for convenience of description and simplicity of description, but do not indicate or imply that the device or element being referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, should not be construed as limiting the present invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance. In the description of the present invention, it should be noted that, unless otherwise explicitly specified or limited, the terms "mounted," "connected," and "disposed" are to be construed broadly, e.g., as meaning either a fixed connection, a removable connection, or an integral connection; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art. The following describes an embodiment of the present invention based on its overall structure.
The invention aims at the problem of wavelength conflict of an AWGR-based optical switching data center network, non-polling scheduling is carried out according to the occupation ratio condition of a sending port cache, the problem of data packet contention and the problem of flexible allocated light bandwidth of an optical network data center can be solved through the scheduling scheme, and finally, a special staggered time slot method is used, so that the waiting time delay caused by a request-response process in a scheduling strategy is avoided.
The AWGR-based data center network architecture mainly comprises a scheduler, a top of rack switch (TOR), an array waveguide grating AWGR and a server. As shown in fig. 2, each rack is internally composed of K servers and a top rack switch (ToR), and the K servers realize communication with other servers in the rack or servers outside the rack through the top rack switch (ToR). As shown in fig. 3This is an inter-chassis data exchange topology, i.e. the core part of the present invention, which is composed of N chassis, a scheduler and a certain number of Arrayed Waveguide Gratings (AWGR). Fig. 4 shows an Arrayed Waveguide Grating (AWGR), which is a passive optical device featuring cyclic wavelength routing and can implement an optical interconnection network. Array Waveguide Gratings (AWGR) are commonly used in conjunction with tunable lasers or fixed wavelength lasers, and the present invention uses a combination of fixed wavelength lasers to implement all-optical switching techniques. The combination of an arrayed waveguide grating and a fixed wavelength laser has the advantages of high capacity, low processing delay and low insertion loss. FIG. 4 is a schematic diagram of a 4x4 Arrayed Waveguide Grating (AWGR) structure with 4 input ports inputting different wavelength signals, wherein
λ in (b) indicates that the signal is an optical signal having a specific wavelength, and the upper corner i is the input port number of the signal, and the lower corner w is the wavelength number of the signal. An Arrayed Waveguide Grating (AWGR) routes the optical signals to corresponding output ports in a cyclic wavelength routing manner.
The top of rack switch (ToR) is implemented by using an FPGA, and its internal module is as shown in fig. 5, where data to be forwarded to other servers in the rack is first uploaded from the servers in the rack to the ethernet switching module, the ethernet switching module distributes the data according to the destination of the data packet, and if the destination is another server in the rack, the data is directly forwarded, and if the destination is another server outside the rack, the data is uploaded to a corresponding cache of the sending module (Tx). The buffers in the top-of-rack switch (ToR) that store the different destination data are grouped and they share one transmit module (Tx) for transmission, but cannot transmit at the same time. Streams sent to different destinations are sent in different time slots through staggered time slots. The function of the transmitting module (Tx) is responsible for transmitting the data packets in the buffer in the form of optical signals in the form of wavelengths, the number of the transmitting modules (Tx) in each top-of-rack switch (ToR) is fixed, and each transmitting module (Tx) can only transmit one different wavelength λ, because the Arrayed Waveguide Grating (AWGR) is used as an optical switching node, which output port the wavelength of each input port can reach is fixed, that is, the route of each flow transmitted in the form of optical wavelengths is fixed. When data is forwarded at a top of rack switch (ToR), the data is forwarded according to a routing table configured in advance. The method comprises the steps that a scheduler collects flow information stored by a cache module in each top of rack switch (TOR) when the scheduling starts, the scheduler sorts all the flow information and analyzes and judges the flow in sequence, if the flow can be sent, the flow is added into a result array, if the flow cannot be sent, the next flow is analyzed and judged, when the result array is full, the flow is analyzed and judged, and finally a control command is issued to each top of rack switch (TOR) according to the information of the result array. And the top rack switch (ToR) controls the corresponding buffer in the transmitting module (Tx) to transmit data according to the command after receiving the command corresponding to the scheduler. Of course, the top-of-rack switch (ToR) will also have an equal number of receive modules (Rx) as transmit modules (Tx).
Wherein the number of wavelengths = number of transmit modules (Tx) = number of top of rack switches (tors)/number of time slots; number of buffers = number of top of rack switches (tors) per number of wavelengths;
the number of Arrayed Waveguide Gratings (AWGR) = number of buffers; arrayed waveguide grating (AWGR port number) = number of wavelengths;
the flow chart of the algorithm part of the invention is shown in fig. 1, the scheduling mechanism is mainly completed by a scheduler realized by an FPGA, and the specifically implemented scheduling algorithm flow is as follows:
a top of rack switch (ToR) will collect the occupation status of each cache in each group and the source node and the destination node corresponding to the cache. The top of rack switch (ToR) will then send a request to the scheduler and wait for a receive command for data transmission. The scheduler receives requests from the top of rack switches (tors) and parses traffic information for subsequent operations.
The number of top-of-rack switches (tors) is set to be N, and each top-of-rack switch (ToR) is provided with M transmitting modules (Tx) (also provided with M receiving modules), so that the number of streams capable of being simultaneously transmitted is at most M × N. The scheduler stores the statistical information of the flows through a priority queue, and the priority queue is sorted according to the flow size of each flow, and the flow size is in front.
Thus, we traverse the elements of the priority queue in order and through the two dictionary array maps T And map R Saving the state of transceivers in each top-of-rack switch (ToR), e.g. map T [1]Indicating the status of M transmit modules (Tx) in a top-of-rack switch (TOR) 1, if map T [1][Tx]If the value of (1) indicates that the transmission module (Tx) is not currently occupied, the stream that needs to be transmitted by the transmission module (Tx) in the next time slot in the priority queue cannot be transmitted, and the transmission of other time slots needs to be waited. map R [1]Status of M receive modules (Rx) in Top of Rack switch (TOR) 1 if map R [1][Rx]If the value of (b) is 0, it means that the receiving module (Rx) is not occupied, and if the value of (b) is 1, it means that it is occupied, the stream which needs to be received by the receiving module (Rx) at the back of the priority queue cannot be transmitted in this time slot, and needs to wait for other time slots to transmit.
Therefore, the factor for deciding whether each stream can be transmitted is the states of the transmitting module (Tx) and the receiving module (Rx), and a sufficient requirement for a stream to be transmitted is that the transmitting module (Tx) and the receiving module (Rx) which it needs to use are simultaneously idle.
The flow which can be sent is put into a result array, if the length of the array reaches M × N, the calculation can be stopped when the data flow which can be sent reaches the upper limit of the number, or the sending number of the flow is not enough, the M × N priority queue is empty, and the calculation is finished.
And finally, the scheduler issues a control instruction to a top of rack switch (ToR) according to the calculation result, the top of rack switch (ToR) judges which stream in the cache can be sent according to the instruction, and then the corresponding cache is opened to send data.
The issuing rule at the beginning of each time slot is calculated at the end of the last time slot. As shown in fig. 6, data transmission starts at time t1 of time slot N, data transmission ends at time t2, a buffer module of a top of rack switch (ToR) collects buffered data and transmits a request to a scheduler, and the scheduler completes at time t3And scheduling and sending a transmission command to each top of rack switch (TOR), and performing data transmission of N +1 time slots at the time t 3. That is to say, network traffic information starts to be collected for scheduling when each time slot is close to the end, because the scheduling time is very short compared with the length of the time slot, the traffic information change generated in the scheduling time can be ignored, it is assumed that 40 servers are arranged in each rack, the bandwidth of each server is 10Gbps, because the time spent for scheduling is ns level, and assuming that the scheduling time is 5ns, the maximum generated data volume is 40x10 9 x5×10 -9 The method is characterized in that the data are transmitted to the same top-of-rack switch (TOR) in the cluster in an extreme condition, the data volume in a buffer is negligible compared with the buffer length, information is collected in advance for calculation, then a new control instruction is issued to the top-of-rack switch (TOR) between the beginning of the next time slot, and the time slot staggered scheduling method can also ensure the scheduling instantaneity.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof, and it is therefore desired that the present embodiments be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein, without any reference thereto being construed as limiting the claim concerned.
Claims (9)
1. A resource scheduling method facing an AWGR optical switching data center network is characterized by comprising the following steps:
s1, a top rack switch collects the occupation condition of each cache of each group and a source node and a destination node corresponding to the cache, then sends a request to a scheduler and waits for receiving a replied command, and the scheduler receives the request sent by each top rack switch and analyzes flow information;
s2, the scheduler stores the counted flow information through a priority queue, the priority queue sorts according to the flow of each flow, and the flow is arranged in front of the priority queue;
s3, traversing the elements of the priority queue in sequence, analyzing and judging the stream, determining whether the current stream can be sent, if so, performing the next step, otherwise, re-performing the S3;
s4, adding the current flow into the result array, stopping calculation if the length of the result array reaches the upper limit of the number or the priority queue is empty, and otherwise, re-executing S3;
and S5, the scheduler issues control commands to each top rack switch according to the information of the result array, and the top rack switch controls corresponding cache in the sending module to send data in a staggered time slot mode after receiving the corresponding commands of the scheduler.
2. The resource scheduling method for the AWGR-oriented optical switched data center network according to claim 1, wherein different buffers in the overhead switch are grouped, and the buffers share one sending module for sending, so that streams sent to different destinations are sent in different time slots.
3. The resource scheduling method for the AWGR-oriented optical switching data center network according to claim 1, wherein the staggered time slot specifically is: and starting to transmit data at the time t1 of the time slot N, finishing data transmission at the time t2, collecting the cache by a cache module of the top rack switch at the time and transmitting a request to a scheduler, finishing scheduling by the scheduler at the time t3 and transmitting a transmission command to each top rack switch, and transmitting data of the time slot N +1 at the time t3 so as to circulate until finishing.
4. The method for resource scheduling in an AWGR-oriented optical switched data center network as claimed in claim 1, wherein the step S3 further comprises traversing the elements of the priority queue according to the order, and passing through two dictionary arrays map T And map R Storing transmit and receive modules in each top-of-rack switchState of (1), if map T And map R The value of the element in the table is 0, which indicates that the sending module or the receiving module is not occupied at present and can send; if map T Or map R The value of the element in the table is 1, which indicates that the sending module or the receiving module is occupied, and then the sending module and the receiving module cannot send the data, and the number of the sending module and the receiving module is the same.
5. The resource scheduling method for the AWGR-oriented optical switched data center network according to claim 1, wherein the S4 specifically is: if the length of the result array reaches MxN, the data stream which can be sent reaches the upper limit of the number, or the length of the array of the stream is not enough for MxN at the moment, but the priority queue is empty, the calculation is stopped; the number of the top-of-rack switches is N, the number of the sending modules in the top-of-rack switches is M, and the number of streams which can be sent simultaneously is at most M multiplied by N.
6. An architecture using the resource scheduling method for the AWGR-oriented optical switched data center network according to any of claims 1-5, comprising a rack, and a scheduler and an arrayed waveguide grating connected to the rack, wherein the rack comprises a top-of-rack switch and a plurality of servers, and the servers communicate with each other through a data switching topology of the top-of-rack switch.
7. The architecture of claim 6, wherein the arrayed waveguide grating structure is a 4x4 array, the arrayed waveguide grating routes optical signals to corresponding output ports in a cyclic wavelength routing manner, 4 input ports arranged inside the arrayed waveguide grating input signals with different wavelengths, the number of transmission modules in each overhead switch is fixed, each transmission module in the same overhead switch can only transmit a unique wavelength λ different from other transmission modules, the arrayed waveguide grating is used as an optical switching node, so that signals with different wavelengths of each input port can reach the fixed output port, and the transmission module is responsible for transmitting data packets in the cache module in the form of optical signals with different wavelengths;
wherein:
λ in (b) indicates that the signal is an optical signal having a specific wavelength, and the upper corner mark i is the input port number of the signal, and the lower corner mark w is the wavelength number of the signal.
8. The framework of claim 6, wherein the internal structure of the top-of-rack switch includes an ethernet switch module, the ethernet switch module is connected to a buffer module disposed between the sending module and the receiving module, the ethernet switch module is connected to a server via a buffer module, data of the server is uploaded to the ethernet switch module from a server in the rack, the ethernet switch module distributes the data according to a destination of the data packet, if the destination is the server in the rack, the data is directly forwarded to the destination server in the rack, and if the destination is the server outside the rack, the data is uploaded to the buffer module corresponding to the sending module, and a control command is waited for data sending.
9. The architecture of claim 6, wherein the number of transmit modules = number of top-of-rack switches/number of time slots = number of wavelengths;
the number of the cache modules = the number of top switches/number of wavelengths = the number of arrayed waveguide gratings;
the number of wavelengths = the number of arrayed waveguide grating ports.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211520300.3A CN115550281B (en) | 2022-11-30 | 2022-11-30 | Resource scheduling method and architecture for AWGR (optical fiber reinforced plastic) optical switching data center network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211520300.3A CN115550281B (en) | 2022-11-30 | 2022-11-30 | Resource scheduling method and architecture for AWGR (optical fiber reinforced plastic) optical switching data center network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115550281A true CN115550281A (en) | 2022-12-30 |
| CN115550281B CN115550281B (en) | 2023-04-28 |
Family
ID=84721644
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211520300.3A Active CN115550281B (en) | 2022-11-30 | 2022-11-30 | Resource scheduling method and architecture for AWGR (optical fiber reinforced plastic) optical switching data center network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115550281B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103441942A (en) * | 2013-08-26 | 2013-12-11 | 重庆大学 | Data center network system and data communication method based on software definition |
| CN105162721A (en) * | 2015-07-31 | 2015-12-16 | 重庆大学 | All-optical interconnection data center network system based on software defined network and data communication method |
| WO2016045055A1 (en) * | 2014-09-25 | 2016-03-31 | Intel Corporation | Network communications using pooled memory in rack-scale architecture |
| CN106941633A (en) * | 2017-02-20 | 2017-07-11 | 武汉邮电科学研究院 | All-optical switching data center network control system and its implementation based on SDN |
| US20180109857A1 (en) * | 2016-10-17 | 2018-04-19 | Electronics And Telecommunications Research Institute | Packet-based optical signal switching control method and apparatus |
| CN110113271A (en) * | 2019-04-04 | 2019-08-09 | 中国科学院计算技术研究所 | A kind of MPI application acceleration system and method based on photoelectricity hybrid switching network |
| CN113709606A (en) * | 2021-08-31 | 2021-11-26 | 山东大学 | Flexible reconfigurable optical switching node system facing elastic optical network and working method |
| WO2022022271A1 (en) * | 2020-07-29 | 2022-02-03 | 浙江大学 | Unbuffered optical interconnection architecture for data center and method |
| CN114827782A (en) * | 2022-04-25 | 2022-07-29 | 南京航空航天大学 | Flow group scheduling method in photoelectric hybrid data center network |
-
2022
- 2022-11-30 CN CN202211520300.3A patent/CN115550281B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103441942A (en) * | 2013-08-26 | 2013-12-11 | 重庆大学 | Data center network system and data communication method based on software definition |
| WO2016045055A1 (en) * | 2014-09-25 | 2016-03-31 | Intel Corporation | Network communications using pooled memory in rack-scale architecture |
| CN105162721A (en) * | 2015-07-31 | 2015-12-16 | 重庆大学 | All-optical interconnection data center network system based on software defined network and data communication method |
| US20180109857A1 (en) * | 2016-10-17 | 2018-04-19 | Electronics And Telecommunications Research Institute | Packet-based optical signal switching control method and apparatus |
| CN106941633A (en) * | 2017-02-20 | 2017-07-11 | 武汉邮电科学研究院 | All-optical switching data center network control system and its implementation based on SDN |
| CN110113271A (en) * | 2019-04-04 | 2019-08-09 | 中国科学院计算技术研究所 | A kind of MPI application acceleration system and method based on photoelectricity hybrid switching network |
| WO2022022271A1 (en) * | 2020-07-29 | 2022-02-03 | 浙江大学 | Unbuffered optical interconnection architecture for data center and method |
| CN113709606A (en) * | 2021-08-31 | 2021-11-26 | 山东大学 | Flexible reconfigurable optical switching node system facing elastic optical network and working method |
| CN114827782A (en) * | 2022-04-25 | 2022-07-29 | 南京航空航天大学 | Flow group scheduling method in photoelectric hybrid data center network |
Non-Patent Citations (1)
| Title |
|---|
| 刘逢清等: "光电混合数据中心的流量分级分类调度研究", 《光电子• 激光》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115550281B (en) | 2023-04-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ye et al. | DOS: A scalable optical switch for datacenters | |
| US8576839B2 (en) | Cascaded contention-free switch modules with interleaved consolidation units | |
| US7843905B2 (en) | Modular high-capacity switch | |
| US7463641B2 (en) | Flexible bandwidth allocation in high-capacity grooming switches | |
| US20080069125A1 (en) | Means and apparatus for a scalable congestion free switching system with intelligent control | |
| US20090080885A1 (en) | Scheduling method and system for optical burst switched networks | |
| CN105162721A (en) | All-optical interconnection data center network system based on software defined network and data communication method | |
| US8064341B2 (en) | Temporal-spatial burst switching | |
| JP7212796B2 (en) | Method and Apparatus for Transparently Embedding Photonic Switching into Electronic Chassis for Scaling Data Center Cloud Systems | |
| CN100428660C (en) | Optical burst exchange node with internal acceleration | |
| Baziana et al. | Collision-free distributed MAC protocol for passive optical intra-rack data center networks | |
| Chunming et al. | Polymorphic Control for Cost‐Effective Design of Optical Networks | |
| CN101193049B (en) | A core node switching structure for integrated edge node | |
| CN115550281B (en) | Resource scheduling method and architecture for AWGR (optical fiber reinforced plastic) optical switching data center network | |
| Smiljanic | Flexible bandwidth allocation in high-capacity packet switches | |
| Cheng et al. | Centralized control plane for passive optical top-of-rack interconnects in data centers | |
| CN101193050B (en) | A method for data receiving and transmitting of core node switching device in optical sudden network | |
| Wang et al. | Efficient protocols for multimedia streams on WDMA networks | |
| Yang et al. | ABOI: AWGR-Based optical interconnects for single-wavelength and multi-wavelength | |
| Binh et al. | Improved WDM packet switch architectures with output channel grouping | |
| Imran | Design of massive optical interconnection system for data centre networks | |
| CN116346745A (en) | Optical packet switch and optical packet switching method | |
| Okorafor | Design and analysis of a 3-dimensional cluster multicomputer architecture using optical interconnection for petaFLOP computing | |
| Che et al. | Switched optical star-topology network with edge electronic buffering and centralized control | |
| Ye | Parallel Wavelength Routing Optical Interconnects for High-Performance Computing Systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |