CN111614779A - Dynamic adjustment method for optimizing and accelerating micro service chain - Google Patents
Dynamic adjustment method for optimizing and accelerating micro service chain Download PDFInfo
- Publication number
- CN111614779A CN111614779A CN202010466541.9A CN202010466541A CN111614779A CN 111614779 A CN111614779 A CN 111614779A CN 202010466541 A CN202010466541 A CN 202010466541A CN 111614779 A CN111614779 A CN 111614779A
- Authority
- CN
- China
- Prior art keywords
- service
- network
- service chain
- micro
- chain
- 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.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
- H04L67/61—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources taking into account QoS or priority requirements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
- H04L67/63—Routing a service request depending on the request content or context
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention relates to a dynamic adjustment method for optimizing and accelerating a micro service chain. It comprises the following steps: 1) actively analyzing the network resource condition in the current working environment, and recording the requirements of specific network services, including the sensitivity of the services to time delay and the minimum resource requirement; 2) reconstructing a service chain by combining the thought of the micro service to enable the service chain to become a micro service chain with variable granularity; 3) the control center divides the combination of the network resource condition and the network service delay sensitivity into four different conditions according to the related environmental data acquired in the earlier stage, and determines a special micro service chain adjustment strategy for each specific condition; 4) drawing a service graph according to a micro service chain adjustment strategy formulated in the early stage and the principle of reducing time delay and reducing resource overhead; 5) according to the service diagram, a series of operations including splitting, merging and paralleling are carried out on the micro service chain, the optimization of the micro service chain is achieved, and the network service is executed according to the optimized micro service chain.
Description
Technical Field
The invention belongs to the technical field of computer communication, and particularly relates to a dynamic adjustment method for optimizing and accelerating a micro service chain.
Background
Network Function Virtualization (NFV) is a new network architecture, and overcomes the disadvantages of high maintenance cost, difficult upgrading, and low flexibility of the conventional network. The NFV uses virtualization technology to implement software operation of the original hardware part in the network, to implement decoupling of software functions and hardware devices, to become specific Network Functions (NF), and to move the NF from the original dedicated devices to the general devices. With NFV, we can break down a particular network service into many NFs. When a data stream passes through each NF in a given order, a Service Function Chain (SFC) is formed. The emergence of NFV has greatly saved deployment, maintenance and upgrade costs for operators and other users. However, as network services are upgraded and their corresponding SFC lengths are increased, service chain delays tend to increase linearly, which is completely unacceptable for delay-sensitive applications. In addition, performance overhead becomes a concern.
Microservices (Microservices) is a software development technology. In the architecture of the microservice architecture, services are fine-grained and protocols are lightweight. With the continuous increase of the current demands of people and the continuous increase of the scale and the complexity of application programs, one application program cannot be developed by a single team, and the fact that a plurality of teams are developed together brings about the problems of testing and modification. When a certain team needs to modify specific functions, the whole program needs to be reconstructed and tested. The micro-service utilizes the characteristics of decentralization, modularization and expansibility of the micro-service to abstract concrete business logic, extract common business capability and make several public services, thereby realizing independent development of each service and greatly reducing business cost. Each service under the micro-service architecture has the characteristics of high cohesion and low coupling, and independent deployment can be achieved. The micro service chain formed by combining the micro service and the service chain has variable granularity, each network function can be used as a single service or is formed by a plurality of functional service blocks, the rapid deployment can be realized aiming at new requirements, and personalized micro service chain optimization can be carried out aiming at different environments and network services.
In the current SFC optimization research, the splitting and parallelization of NF in SFC are currently focused, and the familiar OpenBox and NFP are all the same. However, the split optimization in the 'one-break' manner is not the optimal solution for the accelerated optimization of the service chain, so how to make different optimization strategies according to different situations to further optimize the service chain, and reducing the time delay and the resource overhead becomes a problem which needs to be researched by us.
Disclosure of Invention
The invention aims to provide a dynamic adjustment method for optimizing and accelerating a microservice chain, aiming at the defects of the prior art. The technical scheme for solving the technical problems of the invention comprises the following steps:
the method comprises the following steps that (1) a system actively analyzes the network resource condition in the current working environment and records the requirements of specific network services, including the sensitivity of the services to time delay and the minimum resource requirement;
step (2) reconstructing a service chain by combining the thought of the micro service to enable the service chain to become a micro service chain with variable granularity;
step (3) the control center divides the combination of the network resource status and the network service delay sensitivity into four different situations according to the related environmental data acquired in the early stage, and determines a special micro service chain adjustment strategy for each specific situation;
step (4) drawing a service graph according to a micro service chain adjustment strategy formulated in the early stage and the principles of reducing time delay and reducing resource overhead;
and (5) according to the service graph, performing a series of operations including splitting, merging and paralleling on the micro service chain to realize the optimization of the micro service chain, and executing the network service according to the optimized micro service chain.
When the step (1) is initiated, the system firstly obtains network performance indexes including bandwidth, throughput, time delay and utilization rate in the current working environment through related programs, comprehensively studies and judges the network resource condition, and then analyzes the related indexes of the network service, namely sensitivity to time delay, bandwidth requirements and the like.
And (2) combining the thought of the micro-services, the system reconstructs the service chain to enable the service chain to become the micro-service chain with variable granularity, and each service under the micro-service architecture has the characteristics of high cohesion and low coupling, and can be independently deployed. The micro service chain formed by the combination of the micro service and the service chain has variable granularity, and each network function can be used as a single service or formed by a plurality of functional service blocks.
In the step (3), based on the objective of reducing overhead and time delay for the micro service chain, according to the network resource status and the network service requirement index obtained in the early stage, we combine the two, and artificially set four specific situations, namely, the situation where the network service is not sensitive to time delay and the network resource is sufficient (hereinafter referred to as "condition"), the situation where the network service is not sensitive to time delay and the network resource is insufficient (hereinafter referred to as "condition"), and the situation where the network service is sensitive to time delay and the network resource is insufficient (hereinafter referred to as "condition"), and set a specific micro service chain optimization strategy for the four situations. For the situation, the essential significance of the service chain optimization is small, so the method does not perform essential operation on the service chain optimization, and the network service is still executed according to the original service chain. And aiming at the situation II, the NF which performs modification operation on the data stream is split, and the split functional block and the NF which is not split and only reads the data stream are performed with parallel operation. And for the third case, after all NF are split, only the doubling operation is executed, network resources are saved by combining the same functional blocks, and meanwhile, the resource overhead is reduced without adopting the doubling operation. And aiming at the situation, dividing all NF into different functional blocks according to the service logic of the NF, combining the same functional blocks, and executing the parallel operation of partial functional blocks after judging the possibility of the parallel operation.
In the step (4), after the situation distinction is successfully made, drawing the micro service chain service graph according to the optimization strategy, starting from the original service chain service graph, performing corresponding operation on the network functions in the micro service chain according to the requirements of the optimization strategy, and changing the logic sequence of part of the network functions.
In the step (5), according to the microservice chain service diagram, each network function is broken up into a plurality of micro function blocks or kept as it is under different conditions, and the micro function blocks are combined, executed in parallel or changed in logic position, so as to reduce the resource overhead and the time delay. The invention has the following beneficial effects: the invention solves the important problem of how to make different optimization strategies according to different conditions in the service chain optimization process, namely a dynamic adjustment method for accelerating the optimization of a micro service chain.
From the current research work, the invention is the only set of practical dynamic optimization method with complete solution, and no other effective solution is published at present.
The invention has the following remarkable advantages: 1) the requirement of reducing the time delay of the service chain is met; 2) the requirement of reducing resource overhead of the network service is met; 3) the problem of how to make different optimization strategies in different working environments is solved; 4) the present invention has proven to be practical for use with practical systems.
Drawings
FIG. 1 is a schematic diagram of a service chain;
FIG. 2 is a flow chart of situation determination;
FIG. 3 is a diagram of a microservice chain optimization service;
fig. 4 is a diagram of micro service chain splitting and merging.
Detailed Description
The invention will be further explained with reference to the accompanying drawings.
As shown in fig. 1-4, a method for dynamically adjusting optimization acceleration of a microservice chain includes the following steps:
the method comprises the following steps that (1) a system actively analyzes the network resource condition in the current working environment and records the requirements of specific network services, including the sensitivity of the services to time delay and the minimum resource requirement;
step (2) reconstructing a service chain by combining the thought of the micro service to enable the service chain to become a micro service chain with variable granularity;
step (3) the control center divides the combination of the network resource status and the network service delay sensitivity into four different situations according to the related environmental data acquired in the early stage, and determines a special micro service chain adjustment strategy for each specific situation;
step (4) drawing a service graph according to a micro service chain adjustment strategy formulated in the early stage and the principles of reducing time delay and reducing resource overhead;
and (5) according to the service graph, performing a series of operations including splitting, merging and paralleling on the micro service chain to realize the optimization of the micro service chain, and executing the network service according to the optimized micro service chain.
When the step (1) is initiated, the system firstly obtains network performance indexes including bandwidth, throughput, time delay and utilization rate in the current working environment through related programs, comprehensively studies and judges the network resource condition, and then analyzes the related indexes of the network service, namely sensitivity to time delay, bandwidth requirements and the like. Fig. 1 is a schematic diagram of a service chain before optimization, which is derived from an actual use example of a data center of an operation service provider and is sensitive to delay.
And (3) reconstructing a service chain by combining the idea of the micro service in the step (2) to enable the service chain to become a micro service chain with variable granularity. The NF or functional blocks forming the NF are put into a container by utilizing a container technology to form a service.
In the step (3), based on the objective of reducing overhead and time delay for the micro service chain, according to the network resource status and the network service requirement index obtained in the early stage, we combine the two, and artificially set four specific situations, namely, the situation where the network service is not sensitive to time delay and the network resource is sufficient (hereinafter referred to as "condition"), the situation where the network service is not sensitive to time delay and the network resource is insufficient (hereinafter referred to as "condition"), and the situation where the network service is sensitive to time delay and the network resource is insufficient (hereinafter referred to as "condition"), and set a specific micro service chain optimization strategy for the four situations. Fig. 2 shows a specific situation judgment flowchart.
For the situation, the essential significance of the service chain optimization is small, so the method does not perform essential operation on the service chain optimization, and the network service is still executed according to the original service chain. And aiming at the situation II, the NF which performs modification operation on the data stream is split, and the split functional block and the NF which is not split and only reads the data stream are performed with parallel operation. In four basic operations of adding, deleting, modifying and reading data streams, the time delay caused by the reading operation is far less than that of the other three operations. Therefore, the gains brought by splitting and merging the read-only NF are not easy to be sensed, and the corresponding operation of the NF for executing the adding, deleting and modifying can greatly reduce the time delay and is friendly to the time delay sensitive network service. In addition, the parallel operation of the split functional block and the read-only NF can further improve the time delay. And for the third case, after all NF are split, only the doubling operation is executed, network resources are saved by combining the same functional blocks, and meanwhile, the resource overhead is reduced without adopting the doubling operation. And aiming at the situation, dividing all NF into different functional blocks according to the service logic of the NF, combining the same functional blocks, and executing the parallel operation of partial functional blocks after judging the possibility of the parallel operation.
In the step (4), after the situation distinction is successfully made, drawing the micro service chain service graph according to the optimization strategy, starting from the original service chain service graph, performing corresponding operation on the network functions in the micro service chain according to the requirements of the optimization strategy, and changing the logic sequence of part of the network functions.
As shown in fig. 3 and fig. 4, the micro service chain optimization service diagram and the micro service chain splitting and merging schematic diagram are respectively shown, which belong to a situation (r) for optimizing the service chain of fig. 1 when network resources are scarce. The service chain comprises five network functions of a wide area network optimizer, an edge firewall, a monitor, an application firewall and load balancing. Data streams enter from the wide area network, pass through the service chain and reach specific applications in the data center. It can be known from the analysis that, in the above five NFs, the monitor is used to monitor the number and size of the data packets in the data stream, and the data stream will not be subjected to substantial operations such as addition, deletion, and modification, and the existence of the data stream will not affect the normal operation of the subsequent service chain, so that the parallel optimization of the monitor and the firewall can be realized. Then, we divide the NF into different functional blocks according to its service logic, so as to find and merge completely the same functional blocks, such as reading the packet header. And then arranging the function blocks and drawing a final service diagram according to the service diagram after parallel optimization.
In the step (5), according to the microservice chain service diagram, each network function is broken up into a plurality of micro function blocks or kept as it is under different conditions, and the micro function blocks are combined, executed in parallel or changed in logic position, so as to reduce the resource overhead and the time delay.
Claims (7)
1. A dynamic adjustment method for optimizing and accelerating a micro service chain is characterized by comprising the following steps:
1) actively analyzing the network resource condition in the current working environment, and recording the requirements of specific network services, including the sensitivity of the services to time delay and the minimum resource requirement;
2) reconstructing a service chain by combining the thought of the micro service to enable the service chain to become a micro service chain with variable granularity;
3) the control center divides the combination of the network resource condition and the network service delay sensitivity into four different conditions according to the related environmental data acquired in the earlier stage, and determines a special micro service chain adjustment strategy for each specific condition;
4) drawing a service graph according to a micro service chain adjustment strategy formulated in the early stage and the principle of reducing time delay and reducing resource overhead;
5) according to the service diagram, a series of operations including splitting, merging and paralleling are carried out on the micro service chain, the optimization of the micro service chain is achieved, and the network service is executed according to the optimized micro service chain.
2. The method of claim 1, wherein the network resource status in the current working environment is actively analyzed, and the requirements of specific network services are recorded, including sensitivity of the services to delay and minimum resource requirements: through related programs, network performance indexes including bandwidth, throughput, time delay and utilization rate in the current working environment are firstly obtained, network resource conditions are comprehensively researched and judged, and then related indexes of network services, namely sensitivity to time delay, bandwidth requirements and the like are analyzed.
3. The method according to claim 1, wherein the concept of microservice is combined to reconstruct the service chain into a microservice chain with variable granularity: each service under the micro-service architecture has the characteristics of high cohesion and low coupling, and independent deployment can be achieved.
4. The micro service chain formed by the combination of the micro service and the service chain has variable granularity, and each network function can be used as a single service or formed by a plurality of functional service blocks.
5. The method as claimed in claim 1, wherein the control center divides the combination of the network resource status and the network service delay sensitivity into four different cases according to the related environmental data obtained in the previous period, and determines a specific micro service chain adjustment policy for each specific case: based on the goal of reducing the overhead and the time delay of the micro service chain, according to the network resource condition and the network service demand index acquired in the early stage, the two are combined, four specific situations are set artificially, namely that the network service is insensitive to the time delay and the network resource is sufficient, the network service is sensitive to the time delay and the network resource is sufficient, the network service is insensitive to the time delay and the network resource is insufficient, the network service is sensitive to the time delay and the network resource is insufficient, and specific micro service chain optimization strategies are set aiming at the four situations.
6. The method according to claim 1, wherein the service graph is drawn according to the principle of reducing delay and reducing resource overhead according to the micro service chain adjustment strategy established in the previous stage: after situation differentiation is successfully carried out, drawing a micro service chain service graph according to an optimization strategy, starting from an original service chain service graph, carrying out corresponding operation on network functions in the micro service chain according to the requirements of the optimization strategy, and changing the logic sequence of part of the network functions.
7. The method according to claim 1, wherein a series of operations including splitting, merging, and parallelizing the microservice chain are performed according to a service graph to optimize the microservice chain, and the network service performs the following operations according to the optimized microservice chain: according to the micro service chain service diagram, each network function is scattered into a plurality of micro function blocks or kept as the same under different conditions, and the micro function blocks are combined or executed in parallel or the logic positions of the micro function blocks are changed, so that the reduction of resource overhead and time delay is realized.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010466541.9A CN111614779A (en) | 2020-05-28 | 2020-05-28 | Dynamic adjustment method for optimizing and accelerating micro service chain |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010466541.9A CN111614779A (en) | 2020-05-28 | 2020-05-28 | Dynamic adjustment method for optimizing and accelerating micro service chain |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111614779A true CN111614779A (en) | 2020-09-01 |
Family
ID=72201799
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010466541.9A Pending CN111614779A (en) | 2020-05-28 | 2020-05-28 | Dynamic adjustment method for optimizing and accelerating micro service chain |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111614779A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114615338A (en) * | 2022-04-11 | 2022-06-10 | 河海大学 | Micro-service deployment method and device based on layer sharing in edge environment |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107124303A (en) * | 2017-04-19 | 2017-09-01 | 电子科技大学 | The service chaining optimization method of low transmission time delay |
US20180077080A1 (en) * | 2016-09-15 | 2018-03-15 | Ciena Corporation | Systems and methods for adaptive and intelligent network functions virtualization workload placement |
CN107947965A (en) * | 2017-11-07 | 2018-04-20 | 清华大学 | Service chaining compiler |
CN108923961A (en) * | 2018-06-25 | 2018-11-30 | 电子科技大学 | A kind of multiple entry network service function chain optimization method |
CN110505082A (en) * | 2019-07-26 | 2019-11-26 | 国家电网有限公司 | A kind of NFV service chaining mapping method towards cost and QoS |
CN110601994A (en) * | 2019-10-14 | 2019-12-20 | 南京航空航天大学 | Load balancing method for micro-service chain perception in cloud environment |
US20200067800A1 (en) * | 2018-08-21 | 2020-02-27 | Fujitsu Limited | Service function chain mapping using multiple parallel microservices instances |
CN111130908A (en) * | 2019-12-31 | 2020-05-08 | 中信百信银行股份有限公司 | Micro-service dynamic aggregation and splitting system based on calling flow analysis and prediction |
-
2020
- 2020-05-28 CN CN202010466541.9A patent/CN111614779A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20180077080A1 (en) * | 2016-09-15 | 2018-03-15 | Ciena Corporation | Systems and methods for adaptive and intelligent network functions virtualization workload placement |
CN107124303A (en) * | 2017-04-19 | 2017-09-01 | 电子科技大学 | The service chaining optimization method of low transmission time delay |
CN107947965A (en) * | 2017-11-07 | 2018-04-20 | 清华大学 | Service chaining compiler |
CN108923961A (en) * | 2018-06-25 | 2018-11-30 | 电子科技大学 | A kind of multiple entry network service function chain optimization method |
US20200067800A1 (en) * | 2018-08-21 | 2020-02-27 | Fujitsu Limited | Service function chain mapping using multiple parallel microservices instances |
CN110505082A (en) * | 2019-07-26 | 2019-11-26 | 国家电网有限公司 | A kind of NFV service chaining mapping method towards cost and QoS |
CN110601994A (en) * | 2019-10-14 | 2019-12-20 | 南京航空航天大学 | Load balancing method for micro-service chain perception in cloud environment |
CN111130908A (en) * | 2019-12-31 | 2020-05-08 | 中信百信银行股份有限公司 | Micro-service dynamic aggregation and splitting system based on calling flow analysis and prediction |
Non-Patent Citations (1)
Title |
---|
毛润菲: "微服务系统的可靠性分析及关键问题", 《电子技术与软件工程》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114615338A (en) * | 2022-04-11 | 2022-06-10 | 河海大学 | Micro-service deployment method and device based on layer sharing in edge environment |
CN114615338B (en) * | 2022-04-11 | 2023-07-18 | 河海大学 | Micro-service deployment method and device based on layer sharing in edge environment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108092803B (en) | Method for realizing network element level parallelization service function in network function virtualization environment | |
Grosof et al. | SRPT for multiserver systems | |
CN105812276B (en) | Message forwarding adaptive system and method adopting multi-core processor | |
US20230146912A1 (en) | Method, Apparatus, and Computing Device for Constructing Prediction Model, and Storage Medium | |
CN107391317A (en) | A kind of method, apparatus of data recovery, equipment and computer-readable recording medium | |
CN111367630A (en) | Multi-user multi-priority distributed cooperative processing method based on cloud computing | |
CN110233802B (en) | Method for constructing block chain structure with one main chain and multiple side chains | |
CN107947965B (en) | Service chain compiler | |
CN111614779A (en) | Dynamic adjustment method for optimizing and accelerating micro service chain | |
Zhang et al. | The real-time scheduling strategy based on traffic and load balancing in storm | |
CN108280018A (en) | A kind of node workflow communication overhead efficiency analysis optimization method and system | |
CN116418700A (en) | Distributed data capturing method based on DPDK | |
LU505584B1 (en) | Industrial internet data protection system | |
He et al. | Functional split of in-network deep learning for 6G: A feasibility study | |
CN110365551B (en) | Network information acquisition method, device, equipment and medium | |
CN114630391B (en) | Communication method, device and system under network co-establishment sharing scene and sharing base station | |
CN104424101A (en) | Method and equipment for determining performance interference models of programs | |
CN114138811A (en) | Column calculation optimization method based on Spark SQL | |
CN117389486B (en) | Method, computing device and storage medium for real-time processing EPA network data | |
CN114020218A (en) | Mixed repeating data deleting and scheduling method and system | |
CN113079067A (en) | Union chain performance bottleneck testing system and method based on multidimensional data integration analysis | |
CN113395183A (en) | Virtual node scheduling method and system for network simulation platform VLAN interconnection | |
US10439736B2 (en) | Optical switch control method, and apparatus | |
CN107085536B (en) | Task management method and device | |
CN113419777B (en) | Instruction block conversion execution method and system of programmable switch |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200901 |