Container Image Layer Segmentation for Edge Deployment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In network-constrained environments, such as edge clouds, the transmission of container images is burdensome due to their large size, leading to increased latency in launching applications.

Innovation Solution

A container image format is introduced that defines two types of layers: supplemental and common. Supplemental layers contain computed identifiers associated with each file, while common layers contain real file content. This format allows for efficient reuse of cached layers and reduces the amount of data needed for transmission.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If container images are transmitted in network-constrained environments, then applications can be deployed, but bandwidth usage increases and latency increases

Engineering Contradiction:
Improveapplication deployment speedVSAvoiddata transmission volume
Core Design Contradiction:
ProductivityVSQuantity of substance

Solution Approach 1:

The container image is segmented into multiple layers (common layer and supplemental layer). The common layer containing shared files is cached locally, while only the smaller supplemental layer needs to be transmitted. This segmentation reduces the amount of data that must be transmitted over the network while still enabling complete application deployment.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The common layer is prepared and cached in advance before the actual deployment operation. By pre-caching the bulk of the container image data locally, the system eliminates the need to transmit large volumes of data during the deployment operation itself, significantly reducing latency and bandwidth usage.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If container images are transmitted in network-constrained environments, then applications can be deployed, but transmission time increases

Engineering Contradiction:
Improveapplication launch rateVSAvoidimage transmission time
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The common layer is extracted and separated from the supplemental layer. By taking out the frequently reused common layer and caching it locally, the system eliminates the need to re-transmit this data for subsequent container deployments, dramatically reducing transmission time and improving application launch rates.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The common layer is cached in advance before deployment operations are needed. This preliminary action ensures that when applications need to be deployed, the system only needs to retrieve and process the small supplemental layer rather than transmitting the entire container image, thus reducing transmission time.

Inventive Principle:
Principle #10Preliminary action

3Quantity of substance

If layer caching is implemented, then bandwidth usage is reduced, but system complexity increases

Engineering Contradiction:
Improvenetwork bandwidth consumptionVSAvoidcontainer image format complexity
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The container image format is segmented into two distinct layers with clearly defined roles. The common layer is marked with a specific identifier indicating it can be shared, while the supplemental layer contains application-specific files. This structured segmentation enables automatic layer caching and reuse without requiring complex management logic.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The common layer serves multiple purposes: it contains shared files that can be reused across different container images, acts as a cacheable resource, and provides a foundation for multiple applications. This multi-functionality justifies the added format complexity by delivering significant bandwidth savings across multiple deployments.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP4506843A1Methods and apparatus for container deployment in a network-constrained environment
Publication Date: 2025.02.12 INTEL CORP
  • EP4506843A1 patent drawingFigure 1
  • EP4506843A1 patent drawingFigure 2
  • EP4506843A1 patent drawingFigure 3

AI summary

Systems, apparatus, articles of manufacture, and methods are disclosed for deployment of a container in a network-constrained environment. An example apparatus includes interface circuitry, machine readable instructions, and programmable circuitry to at least one of instantiate or execute the machine readable instructions to access a first unencrypted shared file, access a second encrypted file, create a third file by integrating contents of the first file and the second file, the third file being a Virtual Execution Environment (VEE) image, store the third file, and deploy the third file to a container runtime environment for execution.