feat(coordinator): artifact upload/download endpoints (CTX-05, part 2)

Wire the artifact storage foundation to HTTP.

- PUT /tasks/{id}/artifacts/{filename}: a worker streams a partial result;
  the coordinator verifies lease ownership (foreign worker → 409), streams
  the bytes to blob storage while hashing, and records the metadata. An
  orphaned blob from a failed metadata insert is cleaned up.
- GET /artifacts/{id}/download: streams an artifact back with its content
  type, length, and checksum.
- Ownership is read with a new non-locking TaskRepository.Get, so no row lock
  is held across a long upload. Identity travels in X-Worker-ID / X-Task-Attempt
  headers per the contract; upload/download bypass the short request timeout.
- docker-compose mounts ./data for durable artifact storage; smoke and
  requests.http exercise an upload → foreign-409 → download round-trip.
This commit is contained in:
Efremenko Arhip
2026-07-23 14:12:18 +03:00
parent dbf578c500
commit 6d45406ee0
16 changed files with 276 additions and 24 deletions
+80
View File
@@ -0,0 +1,80 @@
package usecase
import (
"context"
"io"
"github.com/google/uuid"
"github.com/emil28092005/SciMesh/coordinator/internal/domain"
)
// UploadArtifact stores a worker's partial-result bytes and records the metadata.
type UploadArtifact struct {
tasks TaskRepository
artifacts ArtifactRepository
blobs BlobStore
clk Clock
}
func NewUploadArtifact(tasks TaskRepository, artifacts ArtifactRepository,
blobs BlobStore, clk Clock) *UploadArtifact {
return &UploadArtifact{tasks: tasks, artifacts: artifacts, blobs: blobs, clk: clk}
}
func (uc *UploadArtifact) Execute(ctx context.Context, in UploadArtifactInput) (*domain.Artifact, error) {
task, err := uc.tasks.Get(ctx, in.TaskID)
if err != nil {
return nil, err
}
// Only the worker holding the current lease at this attempt may upload the
// task's output — the coordinator never trusts an ownership claim on faith.
if !task.IsLeaseHeldBy(in.WorkerID, in.Attempt) {
return nil, domain.ErrLeaseConflict
}
taskID := task.ID
art, err := domain.NewArtifact(task.JobID, &taskID, domain.ArtifactPartialResult,
in.Filename, in.ContentType, uc.clk.Now())
if err != nil {
return nil, err
}
// Stream to storage first: size and checksum are measured here, by us, not
// taken from the worker. A large shard never sits in memory.
sum, size, err := uc.blobs.Put(ctx, art.StorageKey, in.Body)
if err != nil {
return nil, err
}
art.SetContent(sum, size)
// Persist the record. If that fails the blob would be an orphan, so remove it.
if err := uc.artifacts.Insert(ctx, art); err != nil {
_ = uc.blobs.Delete(ctx, art.StorageKey)
return nil, err
}
return art, nil
}
// DownloadArtifact returns an artifact's metadata together with a reader over
// its bytes. The caller must close the reader.
type DownloadArtifact struct {
artifacts ArtifactRepository
blobs BlobStore
}
func NewDownloadArtifact(artifacts ArtifactRepository, blobs BlobStore) *DownloadArtifact {
return &DownloadArtifact{artifacts: artifacts, blobs: blobs}
}
func (uc *DownloadArtifact) Execute(ctx context.Context, id uuid.UUID) (*domain.Artifact, io.ReadCloser, error) {
a, err := uc.artifacts.Get(ctx, id)
if err != nil {
return nil, nil, err
}
rc, err := uc.blobs.Open(ctx, a.StorageKey)
if err != nil {
return nil, nil, err
}
return a, rc, nil
}
+14 -1
View File
@@ -1,6 +1,10 @@
package usecase
import "github.com/google/uuid"
import (
"io"
"github.com/google/uuid"
)
// Use-case boundary types. Adapters map their wire formats onto these, so the
// HTTP shape can change without touching business code.
@@ -46,6 +50,15 @@ type CompleteTaskInput struct {
Metrics map[string]any
}
type UploadArtifactInput struct {
TaskID uuid.UUID
WorkerID string
Attempt int
Filename string
ContentType string
Body io.Reader
}
type FailTaskInput struct {
TaskID uuid.UUID
WorkerID string
+8
View File
@@ -35,6 +35,11 @@ type TaskRepository interface {
// Returns (nil, nil) when nothing is available.
ClaimNext(ctx context.Context, f ClaimFilter) (*domain.Task, error)
// Get reads a task without locking. Use it for read-only checks (e.g.
// verifying lease ownership before a long upload) where holding a row lock
// across the operation would be wrong.
Get(ctx context.Context, id uuid.UUID) (*domain.Task, error)
// GetForUpdate reads a task and locks its row for the enclosing
// transaction, so read-modify-write use cases stay serialized.
GetForUpdate(ctx context.Context, id uuid.UUID) (*domain.Task, error)
@@ -83,6 +88,9 @@ type ArtifactRepository interface {
type BlobStore interface {
Put(ctx context.Context, key string, r io.Reader) (sha256 string, size int64, err error)
Open(ctx context.Context, key string) (io.ReadCloser, error)
// Delete removes a stored blob. Used to clean up after a metadata insert
// fails, so a committed blob never outlives its (absent) record.
Delete(ctx context.Context, key string) error
}
// TxManager runs a function inside one database transaction. The transaction