Analyses

Analyses are distributed computations within a project. They define the code, configuration, and target nodes for a federated computation.

Creating an Analysis

An analysis requires:

• A parent project (must be approved)
• A master image — the base Docker image for execution
• Analysis code — uploaded to a storage bucket and built into a container

Naming

Every analysis has two name fields:

• name — a URL-friendly identifier (lowercase letters, digits and the characters -, _, .; no whitespace). It is required. If you do not provide one when creating an analysis, a readable name is generated automatically (e.g. brave-otter-1a2b3c). The create form also pre-fills an editable suggestion.
• display_name — an optional, free-form human-readable label. When set, it is shown in the UI in place of the name.

The opaque analysis id (UUID) is no longer surfaced in list views or page headlines — analyses are identified by their display_name (falling back to name).

Node Assignment

When an analysis is created, every approved node of its parent project is automatically assigned to it (as an AnalysisNode). Project nodes that are still pending project approval are skipped — once approved, they can be assigned manually.

Auto-assignment does not bypass the per-analysis approval step: each assigned node still carries its own approval decision. Aggregator nodes are approved automatically, while default nodes start pending and must be approved by the node's realm authority before the analysis can be locked.

You can still add or remove nodes afterwards. Re-assigning a node that is already attached to the analysis updates the existing assignment instead of failing, so manual selection and auto-assignment coexist safely.

Execution Flow

Once approved, an analysis goes through:

1. Configuration — storage buckets are provisioned, configuration is locked
2. Build — the worker builds a Docker image from the analysis code and master image
3. Distribute — the container image is pushed to target node registries
4. Execute — nodes pull and run the container against their local data
5. Complete — results are collected and stored

API Endpoints

Method	Endpoint	Description
GET	/analyses	List analyses
POST	/analyses	Create analysis
GET	/analyses/:id	Get analysis details
PUT	/analyses/:id	Update analysis
DELETE	/analyses/:id	Delete analysis
GET	/analysis-buckets	List analysis storage buckets
GET	/analysis-bucket-files	List analysis bucket files

Analysis Buckets

Each analysis can have associated storage buckets for input data, code, and results. Buckets are managed by the Storage service and provisioned automatically during the configuration phase.