CN111404703B - Time delay optimization method and device, equipment and storage medium - Google Patents
Time delay optimization method and device, equipment and storage medium Download PDFInfo
- Publication number
- CN111404703B CN111404703B CN201910000433.XA CN201910000433A CN111404703B CN 111404703 B CN111404703 B CN 111404703B CN 201910000433 A CN201910000433 A CN 201910000433A CN 111404703 B CN111404703 B CN 111404703B
- Authority
- CN
- China
- Prior art keywords
- server
- temperature
- change
- delay
- simulated annealing
- 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.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/12—Discovery or management of network topologies
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/142—Network analysis or design using statistical or mathematical methods
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L43/00—Arrangements for monitoring or testing data switching networks
- H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
- H04L43/0852—Delays
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Algebra (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Mathematical Physics (AREA)
- Probability & Statistics with Applications (AREA)
- Pure & Applied Mathematics (AREA)
- Environmental & Geological Engineering (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The embodiment of the application discloses a time delay optimization method, a time delay optimization device and a storage medium, wherein the method comprises the following steps: changing the position of a server in a communication network system at a set initial optimization temperature, and determining whether to accept the position change of the server or not through a simulated annealing algorithm; determining an updated temperature according to a simulated annealing algorithm and a previous temperature, wherein the previous temperature comprises the initial optimized temperature; and changing the position of a server in the communication network system at the updated temperature, and determining whether to accept the position change of the server or not through a simulated annealing algorithm until a time delay optimization termination condition is reached.
Description
Technical Field
Embodiments of the present application relate to, but are not limited to, communication technologies, and in particular, to a method, an apparatus, a device, and a storage medium for delay optimization.
Background
The prior art has larger randomness when the system servers are arranged, and most of the servers are arranged according to the experience of technicians, and lack of scientificity and systemicity, so that two servers with high communication frequency and tight connection are possibly positioned in two pools, and network time delay in communication is larger; or only the interconnection of partial servers is considered and the overall network delay of the system is ignored. Particularly as the number of servers and computing nodes increases, it becomes more difficult to plan server locations with a sense of feel. With the rise of applications such as service chains, high-performance computing and big data computing, east-west communication among servers in a data center is obviously increased, frequency is improved, and the rise of technologies such as memory computing and the like improves computing efficiency, so that the influence of network time delay on system performance is further increased.
Disclosure of Invention
In view of this, the embodiments of the present application provide a delay optimization method, apparatus, device, and storage medium for solving at least one problem existing in the related art.
The technical scheme of the embodiment of the application is realized as follows:
the embodiment of the application provides a time delay optimization method, which comprises the following steps:
changing the position of a server in a communication network system at a set initial optimization temperature, and determining whether to accept the position change of the server or not through a simulated annealing algorithm;
determining an updated temperature according to a simulated annealing algorithm and a previous temperature, wherein the previous temperature comprises the initial optimized temperature;
and changing the position of a server in the communication network system at the updated temperature, and determining whether to accept the position change of the server or not through a simulated annealing algorithm until a time delay optimization termination condition is reached.
The embodiment of the application provides a time delay optimizing device, which comprises:
a changing unit configured to attempt to change a server location in the communication network system;
and the optimizing unit is configured to determine whether to accept the position change of the server through a simulated annealing algorithm so as to optimize the time delay of the communication network system.
The embodiment of the application provides a computer device, which comprises a memory and a processor, wherein the memory stores a computer program capable of running on the processor, and the processor realizes the steps in the time delay optimization method when executing the program.
The embodiments of the present application provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the delay optimization method described above.
In the embodiment of the application, as long as the relation between the network overall time delay and the server position of a system is determined, the algorithm can be applied to optimize the server position, so that the influence of the network time delay on the overall performance is reduced, the overall performance of the system is effectively improved, and the application range and the applicability are stronger.
Drawings
Fig. 1 is a schematic diagram of a topology of a fat tree network in the related art;
fig. 2A is a schematic flow chart of an implementation of the delay optimization method in the embodiment of the present application;
fig. 2B is a schematic flowchart of an implementation of the delay optimization method in the embodiment of the present application;
fig. 3 is a schematic flow chart of an implementation of the delay optimization method in the embodiment of the present application;
fig. 4 is a schematic flow chart of an implementation of the delay optimization method in the embodiment of the present application;
Fig. 5 is a schematic diagram of a component structure of a delay optimizing apparatus according to an embodiment of the present application;
fig. 6 is a schematic diagram of a component structure of a delay optimizing apparatus according to an embodiment of the present application;
fig. 7 is a schematic diagram of a hardware entity of a computer device in an embodiment of the present application.
Detailed Description
Fat tree network topology is one of the common networking schemes for non-blocking networking in data centers, and fat tree topology is a Clos network from Charles Clos 1953. In order to do strictly non-blocking, it must be satisfied that the basic switching units that make up the entire network provide non-blocking forwarding. Fat tree topology networks are widely used within data centers, particularly within high performance systems. The fat tree network topology for data centers is typically structured in three stages as shown in fig. 1:
a first stage, edge Switches (Edge Switches) 11, one half of which is connected to the server and the other half of which is connected to the aggregation switch; the second stage is an aggregation switch (aggr. Switches) 12, half of which is connected to the edge switch and half to the core switch; the third stage is a Core switch (Core Switches) 13, all ports of which connect aggregation Switches.
The overall performance of a system is obviously affected by network latency, and when the network latency is large, the computational performance is significantly degraded. The network delay is related to the position between the servers. Obviously, the more closely related servers should be placed as much as possible inside the same pool (pod) or even under the same edge switch to reduce network latency on their path as many communications. And as a system, the overall time delay within the data center should be considered and planned as a whole.
The related art mostly does not consider inter-server arrangement under fat tree topology, and has great randomness and randomness during deployment, so that network delay can have a significant influence on system performance. The related technology has larger randomness in the arrangement of the system servers, and most of the related technology is only laid out according to the experience of technicians, so that scientificity and systemicity are lacked. Two servers with high communication frequency and tight connection are possibly caused to be in two pools, so that network time delay in communication is larger; or only the interconnection of partial servers is considered and the overall network delay of the system is ignored. Particularly as the number of servers and computing nodes increases, it becomes more difficult to plan server locations with a sense of feel. With the rise of applications such as service chains, high-performance computing and big data computing, east-west communication among servers in a data center is obviously increased, frequency is improved, and the rise of technologies such as memory computing and the like improves computing efficiency, so that the influence of network time delay on system performance is further increased. At this time, a more systematic and scientific method is needed to plan and integrate the layout of the servers or the computing nodes so as to optimize the overall network delay of the system.
Fat tree network topology is one of the common networking schemes for non-blocking networking in data centers, and in order to be strictly non-blocking, it must be satisfied that the basic switching elements that make up the entire network provide non-blocking forwarding. The overall performance of the fat tree network topology system is obviously affected by network delay, when the network delay is large, the calculation performance is obviously reduced, the network delay has a relation with the positions of servers, the arrangement of the positions among the servers under the fat tree topology at present has great randomness and randomness, and the network delay has more obvious influence on the system performance.
According to the embodiment of the application, the layout of the server or the computing nodes is planned and comprehensively planned so as to optimize the overall network time delay of the system, and the overall network time delay of the system under the lossless three-level fat tree network is optimized based on the simulated annealing algorithm. The simulated annealing algorithm simulates the metal quenching process, and the combination optimization problem is caused to go from a high-energy state chaotic disordered solution to a low-energy state with higher stability at the beginning through a series of operations similar to 'control cooling'; the simulated annealing algorithm is not trapped in the local optimal solution like a hill climbing method, but has a certain probability of accepting variation after the local optimal solution is reached, so that the local optimal solution can be jumped out and the global optimal solution can be more easily reached.
The technical solutions of the present application are further described in detail below with reference to the drawings and examples.
The present embodiment proposes a time delay optimization method, which is applied to a computer device, where the functions implemented by the method may be implemented by a processor in the computer device calling a program code, and of course the program code may be stored in a computer storage medium, where the computer device includes at least a processor and a storage medium.
Fig. 2A is a schematic flow chart of an implementation of a delay optimization method according to an embodiment of the present application, as shown in fig. 2A, where the method includes:
step S211, changing the position of a server in the communication network system at the set initial optimization temperature;
step S212, determining whether to accept the position change of the server or not through a simulated annealing algorithm;
step S213, determining updated temperature according to the temperature passing through the simulated annealing algorithm and the previous temperature, wherein the previous temperature comprises the initial optimized temperature;
step S214, changing the position of the server in the communication network system at the updated temperature, and determining whether to accept the position change of the server at this time or not through a simulated annealing algorithm until the time delay optimization termination condition is reached.
In some embodiments, the determining whether to accept the current change to the location of the server by the simulated annealing algorithm includes:
calculating the change quantity of network time delay before and after changing;
and determining whether to accept the position change of the server or not according to the change quantity by a simulated annealing algorithm.
In some embodiments, the determining whether to accept the current position change to the server according to the change amount by using a simulated annealing algorithm includes:
when the variation is smaller than zero, the network delay is reduced, and the position change of the server is accepted;
and when the variation is greater than or equal to zero, according to a judgment standard: when (when)
Accepting the change of the location of the server, wherein +.>
Is [0,1]Uniform random number between>
And not accepting the position change of the server.
In some embodiments, assuming the initial optimized temperature is T and the annealing factor of the simulated annealing algorithm is α, the determining the updated temperature based on passing the simulated annealing algorithm and a previous temperature, wherein the previous temperature includes the initial optimized temperature comprises: according to the updated temperature T new Calculated according to the following formula: t (T) new ＝αT old Wherein T is old Is the previous temperature.
In some embodiments, the changing services in the communication network systemA device location, comprising: changing the server location in the communication network system according to a preset number of times, wherein the preset number of times m=100 x (k/2) 2 *P s A second time, wherein k represents the number of switch ports in the communication network system, P s Representing the number of pools planned.
In some embodiments, the delay optimization termination condition: maximum system overall delay t in optimization process at continuous N temperatures max And minimum system overall latency t min The phase difference is smaller than a preset threshold, wherein N is an integer greater than or equal to 2.
The embodiment of the present application further provides a time delay optimization method, and fig. 2B is a schematic implementation flow diagram of the time delay optimization method of the embodiment of the present application, as shown in fig. 2B, where the method includes:
step S201, trying to change the position of a server in a communication network system;
in some embodiments, the communication network system may be a fat tree network system, although other network systems are also possible. The server in the communication network system may be two or more servers under the same edge switch, two or more servers under a different edge switch within the same pod, and two or more servers under a different pod.
In consideration of possible construction or other special reasons, or that some servers are deployed in advance and are not easy to move, all servers are classified and marked as a fixed location Server set Fix (Server) and a movable Server set Rem (Server), so that only the movable servers are considered when the location planning is considered, and the fixed location servers only participate in calculation. In some embodiments, the servers in the communication network system include a fixed location server set and a movable server set, and the attempting to change the server location in the communication network system includes: an attempt is made to change the location of servers in the set of movable servers. To reduce complexity, in some embodiments, the attempting to change the location of a server in a set of movable servers includes: an attempt is made to interact with the locations of two servers in the set of movable servers.
Step S202, determining whether to accept the position change of the server or not through a simulated annealing algorithm so as to optimize the time delay of the communication network system.
In some embodiments, the determining whether to accept the current change to the location of the server by the simulated annealing algorithm includes:
Step a1, calculating the change quantity of network time delay before and after changing;
the amount of change is the network delay after the change (the overall network delay of the system) minus the network delay before the change.
And a2, determining whether to accept the position change of the server or not according to the change quantity by a simulated annealing algorithm.
In some embodiments, step a2, the step of determining whether to accept the current position change to the server according to the change amount by using a simulated annealing algorithm includes:
and when the variation is smaller than zero, the network delay is reduced, and the position change of the server is accepted. When the change amount is zero or more, if the server position in the communication network system is considered as a position change or position change, the probability of accepting the change is P r [accept]＝e -△t/T According to the simulated annealing algorithm, when
When receiving the exchange (i.e. receiving the change of the location of the server), wherein +.>
Is [0,1]Uniform random number between>
The exchange is not accepted (i.e., the change in location of the server is not accepted), wherein: p (P) r [accept]The probability of receiving this variation is represented by Δt, which represents the amount of change in the network delay before and after the change, and T, which represents the current temperature.
In some embodiments, the attempting to change the communication network system middle-wareServer location, comprising: attempting to change a server location in a communication network system a preset number of times, wherein the preset number of times m=100 x (k/2) 2 *P s And secondly, k represents the port number of a switch in the communication network system, and Deltat represents the planned pool number.
In the embodiment of the application, the position of a server in a communication network system is tried to be changed, and whether the position change of the server is accepted or not is determined through a simulated annealing algorithm so as to optimize the time delay of the communication network system; therefore, as long as the relation between the network overall time delay and the server position of a system is determined, the algorithm can be applied to optimize the server position, so that the influence of the network time delay on the overall performance is reduced, the overall performance of the system is effectively improved, and the method has a strong application range and applicability.
The present embodiment proposes a time delay optimization method, which is applied to a computer device, where the functions implemented by the method may be implemented by a processor in the computer device calling a program code, and of course the program code may be stored in a computer storage medium, where the computer device includes at least a processor and a storage medium. Fig. 3 is a schematic flow chart of an implementation of a delay optimization method according to an embodiment of the present application, as shown in fig. 3, where the method includes:
Step S301, under the set initial optimization temperature, trying to change the position of a server in the communication network system, and determining whether to accept the position change of the server or not through a simulated annealing algorithm;
here, the determining whether to accept the current position change to the server through the simulated annealing algorithm includes: calculating the variation of the overall network delay before and after the change; and determining whether to accept the position change of the server or not according to the change quantity by a simulated annealing algorithm.
The step of determining whether to accept the position change of the server according to the change amount by using the simulated annealing algorithm comprises the following steps: when the variation is smaller than zero, the overall network delay is reduced, and the position change of the server is accepted; the variation is greater than or equal to zeroWhen the combination is
When receiving the current position change, wherein->
Is [0,1]Uniform random number between>
The current position change is not accepted.
Step S302, determining updated temperature according to the temperature passing through the simulated annealing algorithm and the previous temperature, wherein the previous temperature comprises the initial optimized temperature;
here, assuming that the initial optimized temperature is T and an annealing factor of the simulated annealing algorithm is α, the determining the updated temperature according to the temperature passing through the simulated annealing algorithm and the previous temperature, wherein the previous temperature includes the initial optimized temperature includes: according to the updated temperature T new Calculated according to the following formula:
T new ＝αT old wherein T is old Is the previous temperature.
Step S303, at the updated temperature, trying to change the position of the server in the communication network system, and determining whether to accept the position change of the server or not by using a simulated annealing algorithm until the time delay optimization termination condition is reached.
Here, the delay optimization termination condition: maximum system overall delay t in optimization process at continuous N temperatures max And minimum system overall latency t min The phase difference is smaller than a preset threshold, wherein N is an integer greater than or equal to 2.
In the embodiment, at a set temperature, trying to change the position of a server in a communication network system, calculating the variation of the overall network time delay before and after the change, and determining whether to accept the position change of the server or not according to the variation by a simulated annealing algorithm; determining updated temperature according to the simulated annealing algorithm and the set temperature; and under the updated temperature, trying to change the position of a server in the communication network system, calculating the change quantity of the overall network delay before and after the change, and determining whether to accept the position change of the server or not according to the change quantity by a simulated annealing algorithm until the delay optimization termination condition is reached.
In some embodiments, the attempting to change the server location in the communication network system includes:
attempting to change a server location in a communication network system a preset number of times, wherein the preset number of times m=100 x (k/2) 2 *P s A second time, wherein k represents the number of switch ports in the communication network system, P s Representing the number of pools planned.
The problem to be solved by the method is to provide a scientific and systematic layout of servers so as to optimize the overall delay parameters of the system under the lossless three-level fat tree network in the data center.
The overall network delay is mainly affected by two parameters, namely, the communication delay between servers and the communication frequency between servers, and the position of the servers obviously has a very critical effect on the overall network delay.
Assuming that the forwarding delay average values of all switches are the same and are Vt, for a general shortest path-based routing algorithm, the single unidirectional communication delay of two servers under the same edge switch is Vt; the communication time delay of two servers in the same pod under the other edge switch is 3Vt; the communication time delay between two servers under the different pod is 5Vt, obviously the positions are different, and the network time delay between the servers can be increased or reduced in multiple. The lossless network can not increase the time delay caused by congestion, and the stability of the time delay of the network can be ensured.
In the fat tree network shown in FIG. 1, when the number of switch ports is k, one pod is composed of (k/2) edge switches and (k/2) aggregation switches, and the maximum number of accommodating servers in the pod is (k/2)/(2). The data center can accommodate (k 3/4) servers at maximum, where (k/2) core switches are required. Assuming that the total number of servers in one data center is n, then the total number is equal toLess P is needed min ＝Floor[n/(k/2) 2 ]+1 pod, provided that the pod number P is planned s Greater than P min The placement of the servers can be ensured, and because of the characteristics of an actual system, the denser the placement of the servers is, the lower the overall network time delay of the system is, so that the empty servers are used for filling the positions of redundant planning servers in all planning pod to ensure the subsequent optimizing effect, and the servers can be considered to have no contact and communication with other servers. Meanwhile, considering that construction or other special reasons can exist in practice, or that some servers are deployed in advance and are not easy to move, all the servers are classified and marked as a fixed location Server set Fix (Server) and a movable Server set Rem (Server), so that when the location planning is considered, only the movable servers are considered, and the fixed location servers only participate in calculation.
At this time, a set of server layout combinations is defined as S (server), and server positions are from 1 to m=100 x (k/2) 2 *P s . The relation function t=f (S (server)) of the overall network delay of the system and the server location at this time can be abstracted according to the communication characteristics between the servers of different systems. The function is mainly affected by two factors, one is single communication time delay between two services caused by the inter-server position, and the other is communication frequency between the two servers. Different computational models and systems can be considered, in which the frequency of communication between servers is substantially fixed or can be reduced to an average or typical value, and the single communication delay caused by the location of the servers is a major factor affecting network delay.
The embodiment of the application provides a technical scheme for optimizing a server layout, wherein the server layout is optimized by using a simulated annealing algorithm. This algorithm simulates the process of metal quenching, and through a series of operations similar to "controlled cooling", the combinatorial optimization problem goes from a state of high energy at the beginning, chaotic solution to a state of low energy and higher stability. In addition, the simulated annealing algorithm does not sink into the local optimal solution like a hill climbing method, but has a certain probability of accepting variation after reaching the local optimal solution, so that the local optimal solution can be jumped out and the global optimal solution can be more easily reached. Fig. 4 is a schematic flow chart of an implementation of a delay optimization method according to an embodiment of the present application, as shown in fig. 4, where the method includes:
Step 401: determining a relation function of the overall network time delay of the system and the position of the server;
step 402: determining the conditions of an initialization optimization temperature T, an annealing factor alpha, the number of ports of a switch in a system, the number of pod, the total position number of a server, a fixed server set and the like;
step 403: randomly generating an initial set of server layouts S (server) init;
the initial set of server arrangements should ensure that the server requiring a fixed location is in its fixed location.
Step 404: randomly selecting two server exchange positions, calculating the delay variation delta t of the whole network, and executing each temperature exchange for M times;
here, at each temperature T cur The following operations were repeated: two servers in a movable server set (Rem) are randomly selected, the positions of the two servers are exchanged, and the change delta t of the overall network time delay t before and after the exchange is calculated. If Deltat <0 or e -△t/T >R∈[0,1]Receiving the position exchange; if Deltat>0 and e -△t/T ≤R∈[0,1]The position exchange is abandoned, the current temperature optimization times m=m-1, wherein M represents the current optimization times, and the value range of M is between 1 and M. If m=0, step 405 is entered.
Here, when Δt <0, it indicates that the overall network delay is reduced, and if the network delay performance is optimized, the optimization is accepted; at 0 or more, the system may be further optimized in the future by a poor variation.
According to the thermodynamic boltzmann distribution: at higher temperatures in the initial case, the state of the system tends to be more chaotic and therefore more easily accepts a worse change to get a subsequent optimization in the future. And after the temperature is reduced, the system tends to be in a better state, and the probability that the change of the deterioration subsequently benefits is reduced. Under the Metropolis algorithm model, according to the Boltzmann distribution:
at temperature T, receiveProbability of variation is P r [accept]＝e -△t/T According to the simulated annealing algorithm, when
When receiving the current position exchange, wherein->
Is [0,1]Uniform random number between>
And not accepting the position exchange.
Here, step 404 is repeatedly performed at a temperature T, totaling m=100 x (k/2) 2 *P s And twice.
Wherein k is the number of switch ports, P s For planning POD number, 100 (k/2) 2 *P s The total number of server locations within the planned pod is calculated such that the number of server location exchanges performed at each temperature is proportional to the total number of location plans.
Step 405: and (5) finishing the current temperature optimization and entering the next temperature optimization.
Here, the temperature is lowered and the new temperature is T new ＝αT old Steps 403 and 404 are re-executed at the new temperature.
Step 406: and (5) optimizing is completed.
Here, steps 404 and 405 are repeatedly performed until the delay optimization termination condition is reached.
Wherein, the time delay optimizes the termination condition: i.e. the maximum system overall time delay t in the optimization process at three consecutive temperatures max And minimum system overall latency t min The phase difference is less than 1%.
Step 407: after the optimization is completed, a group of optimized server layout combinations S is determined optimal (server) and the network overall delay parameter t at that time min 。
In order to solve the technical problem, the embodiment of the present application further provides a total delay optimization device for a lossless fat tree network system, and fig. 5 is a schematic structural diagram of the delay optimization device in the embodiment of the present application, as shown in fig. 5, the device 50 includes a system input module 51, a logic optimization module 52, a system output module 53, and a system storage module 54, where:
the system input module 51 receives system related information including the number of switch ports, the switch latency, the total number of servers and the overall system latency versus server location via file or key-in inputs. Besides the initial optimization parameters, the initial optimization parameters including initial optimization temperature, annealing factors and the like are received.
The logic optimization module 52 is a main body of the device, and according to the input system related information and initial parameters, the logic steps shown in fig. 4 are used for continuously and iteratively optimizing the server positions until the time delay optimization termination condition is reached by using a mode of a large number of attempted server exchanges, so as to form a group of optimized server layouts.
The system storage module 54 puts the files and parameters input by the system into a disk for storage, stores information required by calculation into a memory to be called by the logic optimization module, and stores intermediate results and parameters generated by the logic optimization module. And the final optimization result and the optimization process can be output to a system output module for display and output.
The system output module 53 outputs the optimized server position information and the optimized total time delay result of the system through the file and the display screen, and displays the optimized server position information and the optimized total time delay result to a user.
In the embodiment of the application, 1) the method for judging whether the exchange is accepted or not and the method for judging whether the exchange is accepted or not in the test are carried out on the total time delay of the system by attempting to exchange the server position and judging the server position through a simulated annealing algorithm for the lossless three-level fat tree network system, and 2) the method for judging whether the exchange is accepted or not is carried out on the total time delay of the system, and the time delay is optimized and the termination condition is selected.
Compared with the related art, the embodiment of the application has the following technical advantages: the method for optimizing the overall network time delay of the system in the lossless three-level fat tree network is scientific and effective, and can be applied to optimizing the server position by only determining the relation between the overall network time delay of the system and the server position of the system, so that the distribution of a server or a calculation role is guided, the influence of the network time delay on the overall calculation performance is reduced, the overall performance of the system is effectively improved, and the method has a relatively strong application range and applicability.
Based on the foregoing embodiments, the embodiments of the present application provide a delay optimization apparatus, where the apparatus includes units included, and modules included in the units may be implemented by a processor in a computer device; of course, the method can also be realized by a logic circuit; in an implementation, the processor may be a Central Processing Unit (CPU), a Microprocessor (MPU), a Digital Signal Processor (DSP), a Field Programmable Gate Array (FPGA), or the like.
Fig. 6 is a schematic structural diagram of a delay optimizing apparatus according to an embodiment of the present application, as shown in fig. 6, the apparatus 600 includes:
a changing section 601 for changing a server location in the communication network system at a set initial optimization temperature;
a first determining unit 602, configured to determine whether to accept the current position change to the server through a simulated annealing algorithm;
a second determining means 603 for determining an updated temperature based on the simulated annealing algorithm and a previous temperature, wherein the previous temperature comprises the initial optimized temperature;
and the optimizing component 604 is configured to change the server location in the communication network system at the updated temperature, and determine whether to accept the current change of the server location through the simulated annealing algorithm until the time delay optimization termination condition is reached.
In other embodiments, the first determining means includes:
a calculating sub-component for calculating the variation of the network delay before and after the change;
and the determining sub-component is used for determining whether to accept the position change of the server or not according to the change quantity by a simulated annealing algorithm.
In other embodiments, the determining subcomponent is configured to: when the variation is smaller than zero, the network delay is reduced, and the position change of the server is accepted; and when the variation is greater than or equal to zero, according to a judgment standard: when (when)
Accepting the change of the location of the server, wherein +.>
Is [0,1]Uniform random number among them when
And not accepting the position change of the server.
In other embodiments, assuming that the initial optimal temperature is T and the annealing factor of the simulated annealing algorithm is α, the second determining means is configured to:
according to the updated temperature T new Calculated according to the following formula:
T new ＝αT old wherein T is old Is the previous temperature.
In other embodiments, the changing means is for: changing the server location in the communication network system according to a preset number of times, wherein the preset number of times m=100 x (k/2) 2 *P s A second time, wherein k represents the number of switch ports in the communication network system, P s Representing the number of pools planned.
In other embodiments, the delay optimization termination condition: maximum system overall delay t in optimization process at continuous N temperatures max And minimum system overall latency t min The phase difference is smaller than a preset threshold, wherein N is an integer greater than or equal to 2.
The embodiment of the application further provides a time delay optimizing device, which comprises a changing unit and an optimizing unit, wherein:
a changing unit configured to attempt to change a server location in the communication network system;
and the optimizing unit is configured to determine whether to accept the position change of the server through a simulated annealing algorithm so as to optimize the time delay of the communication network system.
In some embodiments, the optimization unit comprises:
a calculation module configured to calculate a variation of the network delay before and after the change;
and the optimization module is configured to determine whether to accept the position change of the server or not according to the change quantity through a simulated annealing algorithm.
In this embodiment, the server in the communication network system includes a fixed location server set and a movable server set, and the apparatus includes a changing unit and an optimizing unit, where:
The changing unit is configured to attempt to change a location of a server in the movable server set.
And the optimizing unit is configured to determine whether to accept the position change of the server through a simulated annealing algorithm so as to optimize the time delay of the communication network system.
In some embodiments, the changing unit is configured to attempt to interact with the locations of two servers in the set of movable servers.
The embodiment of the application provides a time delay optimizing device, which comprises a changing unit, a determining unit and an optimizing unit, wherein:
the changing unit is configured to: at the set initial optimized temperature, attempting to change the position of a server in the communication network system;
the optimizing unit is configured to determine whether to accept the position change of the server or not through a simulated annealing algorithm;
the determining unit is configured to: determining an updated temperature according to a simulated annealing algorithm and a previous temperature, wherein the previous temperature comprises the initial optimized temperature;
the changing unit is further configured to: attempting to change a server location in the communication network system at the updated temperature;
The optimizing unit is further configured to determine whether to accept the position change of the server through a simulated annealing algorithm until a time delay optimizing termination condition is reached.
In some embodiments, the delay optimization termination condition: maximum system overall delay t in optimization process at continuous N temperatures max And minimum system overall latency t min The phase difference is smaller than a preset threshold, wherein N is an integer greater than or equal to 2.
The embodiment of the application provides a time delay optimizing device, which comprises a changing unit, a determining unit and an optimizing unit, wherein the optimizing unit comprises a calculating module and an optimizing module, and the optimizing unit comprises the following components:
the changing unit is configured to: at the set initial optimized temperature, attempting to change the position of a server in the communication network system;
the calculation module is configured to calculate the variation of the network time delay before and after the change at the set initial optimization temperature;
the optimizing module is configured to determine whether to accept the position change of the server or not according to the change quantity by a simulated annealing algorithm at the set initial optimizing temperature.
The determining unit is configured to: determining an updated temperature according to a simulated annealing algorithm and a previous temperature, wherein the previous temperature comprises the initial optimized temperature;
The changing unit is further configured to: attempting to change a server location in the communication network system at the updated temperature;
the calculating module is configured to calculate the variation of the network time delay before and after the change at the updated temperature;
and the optimizing module is configured to determine whether to accept the position change of the server or not according to the change quantity by using a simulated annealing algorithm at the updated temperature until a time delay optimizing termination condition is reached.
In some embodiments, the optimization module is configured to: when the variation is smaller than zero, the network delay is reduced, and the position change of the server is accepted; when (when)
Accepting the change of the server position, wherein ∈>
Is [0,1]Uniform random number between>
The current position change is not accepted.
In some embodiments, assuming that the initial optimal temperature is T and an annealing factor of a simulated annealing algorithm is α, the determining unit is configured to: according to the updated temperature T new Calculated according to the following formula:
T new ＝αT old wherein T is old Is the previous temperature.
In some embodiments, the changing unit is configured to: attempting to change a server location in a communication network system a preset number of times, wherein the preset number of times m=100 x (k/2) 2 *P s A second time, wherein k represents the number of switch ports in the communication network system, P s Representing the number of pools planned.
The description of the apparatus embodiments above is similar to that of the method embodiments above, with similar advantageous effects as the method embodiments. For technical details not disclosed in the device embodiments of the present application, please refer to the description of the method embodiments of the present application for understanding.
It should be noted that, in the embodiment of the present application, if the above-mentioned delay optimization method is implemented in the form of a software functional module, and is sold or used as a separate product, the delay optimization method may also be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the embodiments of the present application may be essentially or partially contributing to the related art, and the computer software product may be stored in a storage medium, and include several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read Only Memory (ROM), a magnetic disk, an optical disk, or other various media capable of storing program codes. Thus, embodiments of the present application are not limited to any specific combination of hardware and software.
Correspondingly, the embodiment of the application provides a computer device, which comprises a memory and a processor, wherein the memory stores a computer program capable of running on the processor, and the processor realizes the steps in the time delay optimization method when executing the program.
The embodiments of the present application provide a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the delay optimization method described above.
It should be noted here that: the description of the storage medium and apparatus embodiments above is similar to that of the method embodiments described above, with similar benefits as the method embodiments. For technical details not disclosed in the embodiments of the storage medium and the apparatus of the present application, please refer to the description of the method embodiments of the present application for understanding.
It should be noted that, fig. 7 is a schematic diagram of a hardware entity of a computer device in the embodiment of the present application, as shown in fig. 7, the hardware entity of the computer device 700 includes: a processor 701, a communication interface 702 and a memory 703, wherein
The processor 701 generally controls the overall operation of the computer device 700.
The memory 703 is configured to store instructions and applications executable by the processor 701, and may also cache data (e.g., image data, audio data, voice communication data, and video communication data) to be processed or processed by various modules in the processor 701 and the computer device 700, which may be implemented by a FLASH memory (FLASH) or a random access memory (Random Access Memory, RAM).
It should be appreciated that reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. It should be understood that, in various embodiments of the present application, the sequence numbers of the foregoing processes do not mean the order of execution, and the order of execution of the processes should be determined by the functions and internal logic thereof, and should not constitute any limitation on the implementation process of the embodiments of the present application. The foregoing embodiment numbers of the present application are merely for describing, and do not represent advantages or disadvantages of the embodiments.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above described device embodiments are only illustrative, e.g. the division of the units is only one logical function division, and there may be other divisions in practice, such as: multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. In addition, the various components shown or discussed may be coupled or directly coupled or communicatively coupled to each other via some interface, whether indirectly coupled or communicatively coupled to devices or units, whether electrically, mechanically, or otherwise.
The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units; can be located in one place or distributed to a plurality of network units; some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present application may be integrated in one processing unit, or each unit may be separately used as one unit, or two or more units may be integrated in one unit; the integrated units may be implemented in hardware or in hardware plus software functional units.
Those of ordinary skill in the art will appreciate that: all or part of the steps for implementing the above method embodiments may be implemented by hardware related to program instructions, and the foregoing program may be stored in a computer readable storage medium, where the program, when executed, performs steps including the above method embodiments; and the aforementioned storage medium includes: a mobile storage device, a Read Only Memory (ROM), a magnetic disk or an optical disk, or the like, which can store program codes.
Alternatively, the integrated units described above may be stored in a computer readable storage medium if implemented in the form of software functional modules and sold or used as a stand-alone product. Based on such understanding, the technical solutions of the embodiments of the present application may be essentially or partially contributing to the related art, and the computer software product may be stored in a storage medium, and include several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the methods described in the embodiments of the present application. And the aforementioned storage medium includes: various media capable of storing program codes, such as a removable storage device, a ROM, a magnetic disk, or an optical disk.
The foregoing is merely an embodiment of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think about changes or substitutions within the technical scope of the present application, and the changes and substitutions are intended to be covered in the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (6)
1. A method of delay optimization, the method comprising:
changing the position of a server in a communication network system at a set initial optimization temperature, and determining whether to accept the position change of the server or not through a simulated annealing algorithm;
determining an updated temperature according to a simulated annealing algorithm and a previous temperature, wherein the previous temperature comprises the initial optimized temperature;
changing the position of a server in the communication network system at the updated temperature, and determining whether to accept the position change of the server or not through a simulated annealing algorithm until a time delay optimization termination condition is reached;
the determining whether to accept the current position change of the server through the simulated annealing algorithm comprises the following steps: calculating the change quantity of network time delay before and after changing; when the variation is smaller than zero, the network delay is reduced, and the position change of the server is accepted; and when the variation is greater than or equal to zero, according to a judgment standard: when (when)
The position change of the server is accepted when +.>
The change of the location of the server is not accepted at this time, wherein +.>
Is [0,1]A uniform random number among the two, delta T represents the variation of the network time delay before and after the change, and T represents the current temperature;
Assuming that the annealing factor of the simulated annealing algorithm is alpha, determining an updated temperature according to the simulated annealing algorithm and a previous temperature, wherein the previous temperature comprises the initial optimized temperature, and the method comprises the following steps:
updated temperature T new Calculated according to the following formula:
T new ＝αT old wherein T is old Is the previous temperature.
2. The method of claim 1, wherein said changing the location of the server in the communication network system comprises:
changing the server location in the communication network system according to a preset number of times, wherein the preset number of times m=100 x (k/2) 2 *P s A second time, wherein k represents the number of switch ports in the communication network system, P s Representing the number of pools planned.
3. The method of claim 1, wherein the delay optimization termination condition: maximum system overall delay t in optimization process at continuous N temperatures max And minimum system overall latency t min The phase difference is smaller than a preset threshold, wherein N is an integer greater than or equal to 2.
4. A delay optimizing apparatus, the apparatus comprising:
a changing means for changing a server location in the communication network system at a set initial optimization temperature;
A first determining unit for determining whether to accept the current position change to the server by a simulated annealing algorithm;
a second determining means for determining an updated temperature based on a temperature that passed through a simulated annealing algorithm and a previous time, wherein the previous time temperature includes the initial optimized temperature;
the optimizing component is used for changing the position of the server in the communication network system at the updated temperature, and determining whether to accept the position change of the server or not through a simulated annealing algorithm until a time delay optimizing termination condition is reached;
wherein the first determining section includes:
a calculating sub-component for calculating the variation of the network delay before and after the change;
a determining sub-component for: when the variation is smaller than zero, the network delay is reduced, and the position change of the server is accepted; and when the variation is greater than or equal to zero, according to a judgment standard: when (when)
The position change of the server is accepted when +.>
The change of the location of the server is not accepted at this time, wherein +.>
Is [0,1]A uniform random number among the two, delta T represents the variation of the network time delay before and after the change, and T represents the current temperature;
Assuming that the annealing factor of the simulated annealing algorithm is α, the second determining means is for:
the updated temperature T is calculated according to the following formula new ：
T new ＝αT old Wherein T is old Is the previous temperature.
5. A computer device comprising a memory and a processor, the memory storing a computer program executable on the processor, characterized in that the processor implements the steps of the delay optimization method of any of claims 1 to 3 when the program is executed.
6. A computer readable storage medium having stored thereon a computer program, which when executed by a processor performs the steps of the delay optimization method of any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910000433.XA CN111404703B (en) | 2019-01-02 | 2019-01-02 | Time delay optimization method and device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910000433.XA CN111404703B (en) | 2019-01-02 | 2019-01-02 | Time delay optimization method and device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111404703A CN111404703A (en) | 2020-07-10 |
| CN111404703B true CN111404703B (en) | 2023-05-09 |
Family
ID=71413090
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910000433.XA Active CN111404703B (en) | 2019-01-02 | 2019-01-02 | Time delay optimization method and device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111404703B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112135303B (en) * | 2020-09-11 | 2022-12-09 | 苏州浪潮智能科技有限公司 | Data center rapid networking method, system, terminal and storage medium |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103491202B (en) * | 2013-09-09 | 2017-01-04 | 中国科学院计算机网络信息中心 | A kind of DNS node site selecting method based on network measure |
| CN108073740B (en) * | 2016-11-17 | 2021-05-28 | 上海复旦微电子集团股份有限公司 | Simulated annealing method for detailed layout of FPGA |
| CN107135545B (en) * | 2017-04-14 | 2019-11-22 | 中国能源建设集团广东省电力设计研究院有限公司 | Spectrum management method and system based on smart grid bilayer cognitive network architecture |
| CN107766135B (en) * | 2017-09-29 | 2021-04-27 | 东南大学 | Task allocation method based on particle swarm optimization and simulated annealing optimization in moving cloud |
| CN108108224B (en) * | 2017-12-05 | 2019-10-01 | 西南交通大学 | Virtual machine placement method in cloud data center based on ant colony optimization algorithm |
-
2019
- 2019-01-02 CN CN201910000433.XA patent/CN111404703B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN111404703A (en) | 2020-07-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Govorskii et al. | Mathematical modeling of inhomogeneous traffic in a heterogeneous integrated corporate data control system | |
| Sakurai et al. | Two-way communication retrial queues with multiple types of outgoing calls | |
| Moscholios et al. | Performance metrics of a multirate resource sharing teletraffic model with finite sources under the threshold and bandwidth reservation policies | |
| CN109359774B (en) | Workshop scheduling optimization method and device and workshop equipment layout | |
| CN111031519B (en) | Terminal access authentication method and device based on edge calculation | |
| Shin et al. | Enhanced partitioning of dnn layers for uploading from mobile devices to edge servers | |
| CN111404703B (en) | Time delay optimization method and device, equipment and storage medium | |
| Głąbowski et al. | Modelling of multiservice switching networks with overflow links for any traffic class | |
| CN117521576A (en) | Computing resource sharing method, device, equipment and medium | |
| KR101509888B1 (en) | Method and apparatus for message spreading in social network | |
| CN117240727A (en) | Communication network switching method and related equipment | |
| CN112559847A (en) | Method and device for searching nodes in network graph, electronic equipment and storage medium | |
| CN116401992A (en) | Node routing path optimization method and device | |
| CN111582456B (en) | Method, apparatus, device and medium for generating network model information | |
| US10999191B2 (en) | Packet broadcasting mechanism for mesh interconnected multi-computers | |
| CN111131032B (en) | Network path optimization method and device, computer equipment and storage medium | |
| CN111224674A (en) | Decoding method, device and decoder of multi-system LDPC code | |
| CN115022337B (en) | Satellite cloud load balancing method and device, electronic equipment and readable storage medium | |
| KR20240048278A (en) | Optimization methods and dependent systems for disk usage of software-defined storage | |
| Dudin et al. | Multi-threshold control of the B M A P/S M/1/K queue with group services | |
| CN111582482B (en) | Method, apparatus, device and medium for generating network model information | |
| CN116318945B (en) | Multi-target service function chain deployment method based on endophytic dynamic defense architecture | |
| CN116149600B (en) | Method, device, equipment and medium for setting logic circuit of multi-constant multiplier | |
| CN113868395A (en) | Multi-round dialogue generation type model establishing method and system, electronic equipment and medium | |
| CN112906171B (en) | Credible collaborative optimization method and simulation platform for comprehensive energy system |
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 |