| `WebClient` | The browser-based user interface for submitting and managing jobs. |
| `API` | The authenticated service interface for job submission, status, cancellation, history, and deletion. |
| `Worker` | A stateless process that executes attempts and reports progress and outcomes. |
| `IdentityProvider` | The external authority for user identity, administrative roles, and storage authorization scopes. |
| `CoordinationStore` | Shared operational state for queued work, in-flight coordination, live progress, cancellation, and deployment-wide concurrency. |
| `DurableStore` | The source of truth for job identity, ownership, configuration, history, and terminal outcomes. |
| `Source` | An HTTP endpoint, compatible object store, or Hugging Face repository from which a job reads. |
| `StorageAdapter` | The service boundary through which workers write to the target object store. |
| `TargetStore` | The configured S3-compatible object store receiving downloaded objects. |
| `JobProtocol` | The TLA+ specification defining distributed job state, coordination semantics, and safety and liveness properties. |
This document defines Remora's product requirements, system boundaries, major components, result representation and access contract, and allocation of responsibilities. It intentionally does not define the distributed job protocol or implementation-level technology choices.
`JobProtocol` defines job state transitions, at-least-once delivery, claims and fencing, retry scheduling, cancellation races, recovery, and the associated safety and liveness properties. Concrete frameworks, libraries, wire formats, and deployment choices are outside the scope of this specification.
Pulling large objects from remote sources into S3-compatible storage is commonly done with one-off CLI tools or scripts. Those approaches are hard to monitor for long transfers, lack shared retry and resumption, offer little cross-user visibility, and do not enforce authorization over target placement.
## Proposed Solution
Remora exposes a browser UI and API for submitting, monitoring, and managing download jobs into a target object store. Supported source modes are HTTP URLs, compatible object stores, and Hugging Face repository snapshots.
Each download is an independent background job with streaming I/O, automatic retry, live progress, and cancellation. A failed attempt resumes when source and destination support safe resumption; otherwise its replacement starts from the beginning. Jobs default to a configured target bucket and allow per-request bucket and prefix overrides subject to authorization and validation. Every successful job publishes one immutable `ResultManifest`, including single-object jobs.
Identity, coordination, and durable state are external services. Multiple API and worker replicas share them for authenticated multi-user operation.
## Architecture
`WebClient` communicates with `API`, which authorizes requests through `IdentityProvider`, records jobs in `DurableStore`, and submits work through `CoordinationStore`. `Worker` replicas consume jobs and stream data from `Source` through `StorageAdapter` into `TargetStore`.
`API` and `Worker` replicas are stateless. `DurableStore` contains authoritative job records and committed result descriptors. `CoordinationStore` contains reconstructible operational state such as queue entries, leases, deployment-wide permits, cancellation notifications, and live progress.
`TargetStore` contains immutable attempt-scoped objects. A `ResultManifest` roots each attempt-scoped aggregate; `DurableStore` contains the manifest descriptor exposed as the `CommittedResult`. `JobProtocol` is the sole specification of attempt authority and publication behavior.
Workers enforce a configurable outbound-source policy for user-supplied locations on every connection, including redirects. Deployments also restrict worker egress as a defense-in-depth boundary. Private or otherwise sensitive destinations are denied by default unless explicitly allowed by deployment policy.
```mermaid
flowchart TD
subgraph Client
WebClient[WebClient]
end
subgraph Identity
IdentityProvider[IdentityProvider]
end
subgraph Service
API[API]
Worker[Worker]
end
subgraph State
CoordinationStore[CoordinationStore]
DurableStore[DurableStore]
end
subgraph Sources
Source[Source]
end
subgraph Target
StorageAdapter[StorageAdapter]
TargetStore[TargetStore]
end
WebClient -- OIDC --> IdentityProvider
WebClient -- API + live updates --> API
IdentityProvider -- Identity and roles --> API
API -- Submit and coordinate --> CoordinationStore
API -- Record and query --> DurableStore
Worker -- Consume and coordinate --> CoordinationStore
- Every successful job is represented by one immutable `ResultManifest`. A single-object transfer uses a one-member manifest.
- A manifest is the exact inventory of immutable member identities and carries sufficient integrity metadata to validate the aggregate.
- Manifest construction uses memory bounded independently of member count and total result size.
- Remora resolves a result only from its `CommittedResult`; it never derives one by listing a target bucket or prefix.
- An authorized reader may access committed members directly in `TargetStore` using exact manifest identities or time-limited URLs.
- Raw target listing is operational storage visibility, not result resolution. Clients with independent target credentials may observe uncommitted attempt objects.
- A stable user-chosen object key or atomically visible raw target prefix is not part of a job result. Providing one requires a separate export contract.
-`DurableStore` retains exact attempt ownership and pending cleanup work so cleanup can recover independently of `CoordinationStore`.
- Ambient cleanup processes are stateless and idempotent. Worker-local cleanup may reduce latency but is not the only cleanup path.
- Cleanup revalidates authority in `DurableStore` and addresses exact attempt object or multipart identities. Missing coordination state, target-object age, and raw prefix membership do not grant cleanup authority.
- Target lifecycle rules are storage hygiene only. They do not determine result publication or cleanup eligibility and must not expire completed attempt objects that Remora may still protect.
- A reusable checkpoint binds an immutable `Source` identity, an exact completed boundary, and validated destination state. If any proof is absent or changed, the replacement restarts that member from the beginning.
- HTTP identity includes the resolved final URL, a strong representation validator as defined by `RFC 9110`, representation length, and identity content encoding. Resume requires a conditional byte-range response whose validator, URL, encoding, and range exactly match the checkpoint. Weak validators, length, or modification time alone do not qualify.
- Object-store identity includes endpoint, bucket, key, size, and an immutable version ID when available. Without a version ID, resume requires a matching conditional range read. Destination multipart reuse additionally requires the same endpoint, bucket, key, upload ID, and exact re-listed completed parts.
- A Hugging Face branch or tag is resolved once to a full commit. The source identity includes repository type and ID, that commit, and the exact file inventory with Git blob or LFS identities. Retries and cache reuse remain pinned to the stored commit and member identities.
- A checkpoint never authorizes best-effort concatenation. An unavailable explicitly pinned or resolved snapshot produces a structured source-changed outcome rather than silently resolving a newer snapshot.
-`JobProtocol` remains the sole owner of execution phase. Deletion never rewrites a terminal execution outcome and is available only after a job is terminal; cancelling active work remains a separate protocol operation.
- Soft deletion tombstones Remora visibility immediately without deleting target objects. It is idempotent and may be restored until hard deletion begins.
- Hard deletion atomically tombstones visibility and creates one durable, idempotent deletion operation. Once that operation exists, the result cannot be restored even while physical deletion is pending or retrying.
- A hard-deletion operation uses the `ResultManifest` and durable attempt ownership to materialize exact deletion work with bounded memory. It removes committed members and eligible attempt data, treats an already-missing exact identity as success, and deletes the manifest only after its members.
- Physical deletion is asynchronous and recoverable across worker loss and transient target failures. It does not depend on client request lifetime or reconstruct work by listing a target prefix.
- A permanent minimal tombstone retains job identity, owner, terminal outcome, deletion identity, and audit timing. Retention of more detailed job metadata is deployment policy and never restores result visibility after hard deletion begins.
| Reliability | Preserve durable authority and committed results across worker or coordination failures; reconcile in-flight work by protocol |
| Observability | Expose job status, progress, and errors through the UI and API |
| Scalability | Support multiple stateless API and worker replicas using shared state services |
| Performance | Stream multi-gigabyte objects without buffering complete objects in worker memory |
| Security | Authenticate and authorize all operations; do not expose long-lived source or target credentials to the UI |
| Deployment | Deploy as a standard Kubernetes service using externally managed identity, state, and object storage |
## Responsibility Boundaries
- This architecture specification defines components, trust boundaries, externally visible behavior, and responsibility allocation.
-`JobProtocol` defines distributed state and coordination semantics, including safety and liveness invariants.
-`JobProtocol` is the normative implementation contract for durable transitions, fencing, leases, retries, cancellation, publication, and reconciliation. Implementations may optimize storage layout but must refine its observable behavior and preserve its linearization points.
