百科.dev
全部条目AI 编程趋势榜开源项目技术资讯提交条目
登录
< 返回工具列表
S

soci-snapshotter

> 编程语言
开源

一个容器快照插件,可让标准的 OCI 映像在不需要构建时转换的情况下被懒加载。

758 stars0 点赞0 次浏览
访问官网GitHub

工具介绍

一个容器快照插件,可让标准的 OCI 映像在不需要构建时转换的情况下被懒加载。

SOCI Snapshotter

SOCI Snapshotter is a containerd snapshotter plugin. It enables standard OCI images to be lazily loaded without requiring a build-time conversion step. "SOCI" is short for "Seekable OCI", and is pronounced "so-CHEE".

The standard method for launching containers starts with a setup phase during which the container image data is completely downloaded from a remote registry and a filesystem is assembled. The application is not launched until this process is complete. Using a representative suite of images, Harter et al FAST '16 found that image download accounts for 76% of container startup time, but on average only 6.4% of the fetched data is actually needed for the container to start doing useful work.

One approach for addressing this is to eliminate the need to download the entire image before launching the container, and to instead lazily load data on demand, and also prefetch data in the background.

Design considerations

No image conversion

NOTE This section describes SOCI Index Manifest v1. While the lack of an image conversion step is appealing, for production scenarios it also creates the potential for performance changes across multiple dimensions if an index is added to or removed from an existing image that is widely deployed. To address these downsides, we introduced SOCI Index Manifest v2 which does use a build-time conversion step. For more information see the SOCI Index Manifest v2 documentation

Existing lazy loading snapshotters rely on a build-time conversion step, to produce a new image artifact. This is problematic for container developers who won't or can't modify their CI/CD pipeline, or don't want to manage the cost and complexity of keeping copies of images in two formats. It also creates problems for image signing, since the conversion step invalidates any signatures that were created against the original OCI image.

SOCI addresses these issues by loading from the original, unmodified OCI image. Instead of converting the image, it builds a separate index artifact (the "SOCI index"), which lives in the remote registry, right next to the image itself. At container launch time, SOCI Snapshotter queries the registry for the presence of the SOCI index using the mechanism developed by the OCI Reference Types working group.

Workload-specific load order optimization

Another big consideration that we haven't implemented/integrated into SOCI is image load order based on your specific workload. See design README for more details.

Documentation

  • Getting Started: walk through SOCI setups and features.
  • Build: how to build SOCI from source, test SOCI (and contribute).
  • Install: how to install SOCI as a systemd unit.
  • Debug and Useful Commands: accessing logs/metrics and debugging common errors.
  • Prefetch: understanding and managing prefetch artifacts (experimental).
  • Glossary: glossary we use in the project.

Integration-specific documentation

  • SOCI on Kubernetes: an overview of how to use SOCI on Kubernetes in general
  • SOCI on Amazon Elastic Kubernetes Service (EKS): a walk through for setting up SOCI on Amazon EKS

Project Origin

There are a few different lazy loading projects in the containerd snapshotter community. This project began as a fork of the popular Stargz-snapshotter project from commit 743e5e70a7fdec9cd4ab218e1d4782fbbd253803 with the intention of an upstream patch. During development the changes were fundamental enough that the decision was made to create soci-snapshotter as a standalone project. Soci-snapshotter builds on stargz's success and innovative ideas. Long term, this project intends and hopes to join containerd as a non-core project and intends to follow CNCF best practices.

Issues· 0 开放

查看全部 Issues在 GitHub 打开

暂无开放 Issues,或尚未同步最近议题。

> 标签

Go

暂无评论,来聊聊你的看法吧

> 工具信息

发布日期2026年8月1日
最后更新2026年9月17日
分类编程语言
定价开源

> 相关工具

T
TypeScript
JavaScript 的超集,为前端与全栈提供静态类型
P
Python
通用编程语言,广泛用于 Web、数据与 AI
G
Go
Google 推出的简洁高效系统语言