device-mapper-utils/blk-stash
在 GitHub 查看Remote repositories (send/receive/list archive support)
Open
#16 创建于 2025年7月23日
enhancementhelp wanted
仓库指标
- Star
- (0 star)
- PR 合并指标
- (PR 指标待抓取)
描述
Remote repositories
Summary
Support for remote repositories (e.g., storing or retrieving files/block devices over SSH, S3, or HTTP) is a planned feature of blk-archive, but not yet implemented. Before committing to a specific milestone, we need alpha release feedback to assess how urgent this feature is for real-world use cases.
The design for remote repository support should be finalized, but actual implementation may be postponed depending on feedback.
📌 Related design work is being tracked in PR #1.
Goals
- Gather feedback from alpha users on whether remote repository support is a blocking need
- Finalize the design for how remote repositories will be configured, accessed, and authenticated
- Define protocol, fallback behavior, error handling, and performance considerations
- Make implementation timeline dependent on urgency expressed by users
Tasks
- Draft and review design for remote repository support
- Document intended remote transport options (e.g., SSH, S3, HTTP, custom protocols)
- Determine how to configure remotes (e.g., CLI flags, config files)
- Identify expected use cases (e.g., backups over WAN, remote de-duplication)
- Get feedback from early adopters on urgency and blockers
- Defer implementation to a later release if demand is not warranted
Questions for Users
- Is remote support needed now, or can it wait until a later release?
- What protocols do you expect to use (e.g., SSL, ssh tunnel, object storage)?
- Do you expect full remote access (read + write), or read-only (e.g., for restoration)?
Design Considerations
- Bandwidth efficiency
- Authentication and encryption
- Retry behavior, resume support
- Read-only vs. read-write remotes
- Compatibility with local caching
Benefits
- Enables distributed or cloud-based archive storage
- Supports multi-location de-duplication workflows
- Improves flexibility for backups, archiving, and disaster recovery
This is a large feature and will likely need multiple sub-issues for full implementation