apache / apache/texera

Curated computing-unit images

Open
#8,466 3 comments 0 reactions 1 assignee Claimed by @tanishqgandhi1908 View on GitHub
Dominant language
Scala
Stars
314
Forks
187
Avg merge
1d 21h
Merged PRs (30d)
214

Description

### Feature Summary

Every computing unit runs the same image, fixed when the cluster is installed. Work needing a different Python version, a system package, or a library built from source cannot run, because a Python virtual environment only holds pip packages. AlphaFold 3 fails all three ways at once.

This lets an administrator register an image from a public registry. It then appears in a dropdown when creating a computing unit, and that unit runs it. Nobody else is affected.

### Proposed Solution or Design

An administrator pastes a reference — `owner/name:tag`, or a Docker Hub page address — on an admin page. Texera checks the image starts `computing-unit-master`, which means it was built FROM the Texera computing-unit image, and records the digest the reference points at. A user then picks it when creating a unit.

- Nothing is copied and no registry is added. Units pull the reference the same way the default computing-unit image is already pulled.
- Pinned to the digest, so a tag its owner moves later cannot change what runs.
- Checked when registered, in front of the administrator, not when a user's unit fails to start.
- Chosen when the unit is created and fixed for its lifetime.

Out of scope: building images, private registries, per-operator images.

Discussion: #7812

### Affected Area

Deployment / Infrastructure, Workflow UI

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.