CN112732411A - Cloud-side-architecture-based method for issuing cloud-side inference model to side - Google Patents
Cloud-side-architecture-based method for issuing cloud-side inference model to side Download PDFInfo
- Publication number
- CN112732411A CN112732411A CN202110108859.4A CN202110108859A CN112732411A CN 112732411 A CN112732411 A CN 112732411A CN 202110108859 A CN202110108859 A CN 202110108859A CN 112732411 A CN112732411 A CN 112732411A
- Authority
- CN
- China
- Prior art keywords
- model
- cloud
- container
- data volume
- edge
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
- G06F9/5072—Grid computing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/45562—Creating, deleting, cloning virtual machine instances
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/4557—Distribution of virtual machine instances; Migration and load balancing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/45595—Network integration; Enabling network access in virtual machine instances
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2209/00—Indexing scheme relating to G06F9/00
- G06F2209/50—Indexing scheme relating to G06F9/50
- G06F2209/502—Proximity
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a method for issuing a cloud end inference model to an edge end based on a cloud edge end architecture. In this architecture, a visualization container management tool portal to the open source is used, as well as a proprietary docker warehouse harbor. The method comprises the steps of registering docker of an edge end to a cloud end, managing the docker of the edge end in a cloud end portal, constructing a trained inference model into a model mirror image, uploading the model mirror image to a hardor warehouse, pulling the inference model mirror image of the hardor warehouse to the edge end through the portal, establishing an edge end model data volume container, associating a path where a model in the inference model container is located with the data volume container, and synchronizing the model in the inference model container to a data volume, wherein the model is successfully issued to an edge end. Other application programs at the side end can be associated with the data volume, and follow-up work is carried out by using the issued inference model.
Description
Technical Field
The invention relates to a cloud model issuing method based on a cloud edge architecture, and particularly relates to the field of edge computing.
Background
The container has the excellent characteristics of light weight, safety, quick start and the like, and has natural light weight and portability, so that the container is widely applied to the scene of edge calculation. Although cloud computing is more and more mature, some defects are generated when all computing is placed in the cloud, so that an edge computing platform is built, a management method of resources on the cloud is extended to an edge, and resources and equipment of the edge can be managed seamlessly. How to realize the arrangement, deployment and configuration of the cloud to the end application is one of the problems that the edge platform needs to solve. Portainer is a lightweight visual container management tool in which a method of creating a managed edge computing environment for an edge broker is provided, by which the docker environment of the edge environment can be managed. In addition, the development and operation of Docker container applications cannot be reliably managed by mirroring, and although Docker officials provide a public mirror repository, deployment of Registry within a private environment is also necessary from the aspects of security and efficiency. Harbor is an enterprise-level Docker Registry management project sourced by VMware corporation that provides conditions for storage of model images.
In the using process of the Docker, data is required to be persisted, or data sharing is required among a plurality of containers, so that the data operation of the Docker container is involved. When a program runs in a docker container, communication with programs in other containers or programs outside the container is needed, and data exchange is needed. Docker uses a specialized form of data volume to provide a system that facilitates the storage of data. The data volume is independent of the containers, a plurality of containers can share the same data volume, and file data can be shared among the containers through the data volume. To better share data between containers, data volume containers may be used. Through the data volume container, the data volume can be classified and summarized more easily, the relationship between the container and the data volume can be managed better, and the life cycle of the data volume can be controlled more reasonably. The data volume container is a container specially used for storing the data volume, and when the data volume needs to be used in other containers, the data volume is mounted from the data volume container instead of taking the catalog of the host as the data volume. Data volumes exist independently of containers, and data volume containers do not directly manage or control data volumes, but rather are bridges for other containers using data volumes. A data volume container is a bridge that connects other containers to the data volume, and a data volume may exist independently of a container. At present, cloud computing is more and more mature, but some defects can be caused when all computing is placed in a cloud end, so that an edge computing platform is built, a management method of resources on the cloud is extended to an edge end, and resources and equipment of the edge end can be managed seamlessly. How to realize the arrangement, deployment and configuration of the cloud to the end application is one of the problems that the edge platform needs to solve.
Disclosure of Invention
The invention aims to provide a cloud model issuing method based on a cloud edge architecture, which can realize issuing of a cloud operation model, issue the model from a cloud to an edge, and realize loading, deleting, upgrading and the like of the edge model.
In order to achieve the purpose, the invention is realized by the following technical scheme:
a cloud model issuing method based on a cloud edge architecture comprises the following steps:
1) establishing a harbor warehouse at the cloud end;
2) mirror image uploading of the inference model to a hardor;
3) a portal management tool is installed in the cloud;
4) registering the edge terminal to the cloud terminal by utilizing a portal agent;
5) drawing a hardor inference model mirror image by the side end;
6) the data volume container is created by the side end;
7) mounting a data volume container, establishing a mapping relation of model data, and starting an inference model container;
8) and synchronizing the inference model to the edge data volume container, starting the application program container by the edge, mounting the data volume container, synchronizing the inference model to the application program container, and enabling the application program to realize the model inference work.
Preferably, the data volume container in step 8 is a data volume container containing an inference model.
The invention has the advantages that: the cloud operation model can be issued, the model is issued from the cloud to the edge, and loading, deleting, upgrading and the like of the edge model are achieved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention.
Fig. 1 is a schematic front view of an embodiment 1 of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
As shown in fig. 1, the cloud establishes a mirror image private warehouse, and uploads a mirror image of the inference model to the vault; the method comprises the following steps that a portal management tool is installed in a cloud, and an edge is registered to the cloud by using an edge registration method provided by the tool; then issuing the inference model mirror image in the hardor to the edge, starting the inference model mirror image container and binding the data volume container established in advance, and finishing issuing the inference model to the edge; and starting the application container and binding the data volume container containing the inference model, so that the application program can load the inference model to carry out the next inference related work. The method comprises the following specific steps:
1) establishing a harbor warehouse at the cloud end;
2) mirror image uploading of the inference model to a hardor;
3) a portal management tool is installed in the cloud;
4) registering the edge terminal to the cloud terminal by utilizing a portal agent;
5) drawing a hardor inference model mirror image by the side end;
6) the data volume container is created by the side end;
7) mounting a data volume container, establishing a mapping relation of model data, and starting an inference model container;
8) and synchronizing the inference model to the edge data volume container, starting the application program container by the edge, mounting the data volume container containing the inference model, synchronizing the inference model to the application program container, and realizing the model inference work by the application program.
Claims (2)
1. A cloud model issuing method based on a cloud edge architecture is characterized by comprising the following steps:
1) establishing a harbor warehouse at the cloud end;
2) mirror image uploading of the inference model to a hardor;
3) a portal management tool is installed in the cloud;
4) registering the edge terminal to the cloud terminal by utilizing a portal agent;
5) drawing a hardor inference model mirror image by the side end;
6) the data volume container is created by the side end;
7) mounting a data volume container, establishing a mapping relation of model data, and starting an inference model container;
8) and synchronizing the inference model to the edge data volume container, starting the application program container by the edge, mounting the data volume container, synchronizing the inference model to the application program container, and enabling the application program to realize the model inference work.
2. The cloud model issuing method based on the cloud edge architecture of claim 1, wherein the data volume container in step 8 is a data volume container containing an inference model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110108859.4A CN112732411A (en) | 2021-01-27 | 2021-01-27 | Cloud-side-architecture-based method for issuing cloud-side inference model to side |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110108859.4A CN112732411A (en) | 2021-01-27 | 2021-01-27 | Cloud-side-architecture-based method for issuing cloud-side inference model to side |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112732411A true CN112732411A (en) | 2021-04-30 |
Family
ID=75594158
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110108859.4A Pending CN112732411A (en) | 2021-01-27 | 2021-01-27 | Cloud-side-architecture-based method for issuing cloud-side inference model to side |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112732411A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114924845A (en) * | 2022-07-21 | 2022-08-19 | 合肥中科类脑智能技术有限公司 | Mirror image delay loading method and system suitable for edge AI scene |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110109779A (en) * | 2019-07-02 | 2019-08-09 | 南京云信达科技有限公司 | A method of data, which are carried out, based on micro services restores rehearsal environmental structure |
| CN111414233A (en) * | 2020-03-20 | 2020-07-14 | 京东数字科技控股有限公司 | Online model reasoning system |
-
2021
- 2021-01-27 CN CN202110108859.4A patent/CN112732411A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110109779A (en) * | 2019-07-02 | 2019-08-09 | 南京云信达科技有限公司 | A method of data, which are carried out, based on micro services restores rehearsal environmental structure |
| CN111414233A (en) * | 2020-03-20 | 2020-07-14 | 京东数字科技控股有限公司 | Online model reasoning system |
Non-Patent Citations (3)
| Title |
|---|
| 三度: "portainer安装，配置自定义镜像仓库拉取镜像", 《博客园，HTTPS://WWW.CNBLOGS.COM/SANDUZXCVBNM/P/13180692.HTML》 * |
| 银河架构师: "docker进阶之路-基础篇│二：portainer安装与使用基础", 《CSDN，HTTPS://BLOG.CSDN.NET/LIUMINGLEI1987/ARTICLE/》 * |
| 风间悠香: "[工具]Docker及Portainer GUI的使用", 《博客园，HTTPS://WWW.CNBLOGS.COM/LEOKALE-ZZ/P/12389912.HTML》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114924845A (en) * | 2022-07-21 | 2022-08-19 | 合肥中科类脑智能技术有限公司 | Mirror image delay loading method and system suitable for edge AI scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107566541A (en) | Capacitor network resource allocation methods, system, storage medium and electronic equipment | |
| CN110995480B (en) | Block chain network deployment method, device, electronic equipment and medium | |
| CN105684357A (en) | Management of addresses in virtual machines | |
| US11121577B2 (en) | Satisfying building energy demand using mobile energy storage | |
| CN110502217B (en) | ROS-based robot cloud platform design method | |
| CN104067257A (en) | Automated event management | |
| CN103036855B (en) | A kind of equipment that realizes of rights management and method | |
| CN104516730A (en) | Data processing method and device | |
| CN114500623B (en) | Network target range interconnection and intercommunication method, device, equipment and readable storage medium | |
| CN105359088A (en) | Method to optimize provisioning time with dynamically generated virtual disk contents | |
| US20200272433A1 (en) | Workflow engine tool | |
| CN103685608A (en) | Method and device for automatically configuring IP (Internet Protocol) address of security virtual machine | |
| CN109802934A (en) | A kind of MEC system based on container cloud platform | |
| CN105786593A (en) | Method and system for re-launching contextually related application sets | |
| CN103036916A (en) | Method, device and system thereof for calling remote hardware resources | |
| US11954064B2 (en) | Framework for migrating applications to cloud computing environment | |
| CN112732411A (en) | Cloud-side-architecture-based method for issuing cloud-side inference model to side | |
| CN113626002A (en) | Service execution method and device | |
| CN117112122A (en) | Cluster deployment method and device | |
| CN104951238A (en) | Method and device for managing data storage in distributed virtual environment | |
| CN104408580A (en) | Light weight workflow engine for countersignature platform and realization method of light weight workflow engine | |
| CN105447033A (en) | Method and apparatus for generating initial copy in replication initialization | |
| CN113362037A (en) | Coal mine intelligent management system and method based on edge cloud | |
| CN104424012A (en) | Method and equipment used for providing user-defined virtual device | |
| CN114363164B (en) | Cloud network service arrangement control method, system, storage medium and electronic equipment |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210430 |