package httpapi import ( "context" "encoding/json" "errors" "net/http" "strings" "time" "knowledge-governance/internal/governance" "knowledge-governance/internal/jobs" ) // 按指定整库版本恢复:与同步分开的作业类型,需要**单独的管理员授权**, // 授权摘要绑定"恢复到这个快照"这一件事,不能拿同步授权顶替。 const libraryRestoreJobType = "library.restore" func libraryRestoreParamsDigest(scope, snapshotID string) (string, error) { return governance.LibraryRequestDigest(scope, governance.LibraryJobRestore, map[string]any{"restore_snapshot_id": snapshotID}) } type libraryRestoreRequest struct { Scope string `json:"scope"` SnapshotID string `json:"snapshot_id"` } func libraryRestoreEnqueue(w http.ResponseWriter, r *http.Request, config Config) { if r.Method != http.MethodPost { writeJSON(w, http.StatusMethodNotAllowed, map[string]string{"error": "method not allowed"}) return } identity, ok := sessionFrom(r.Context()) if !ok || roleRank(identity.Role) < roleRank(roleAdmin) { writeJSON(w, http.StatusForbidden, map[string]any{"error": "按版本恢复仅管理员可发起"}) return } store, ok := libraryStore(config) if !ok { writeJSON(w, http.StatusServiceUnavailable, map[string]any{"code": "library_restore_unavailable"}) return } var request libraryRestoreRequest if err := json.NewDecoder(http.MaxBytesReader(w, r.Body, 4096)).Decode(&request); err != nil { writeJSON(w, http.StatusBadRequest, map[string]string{"error": "invalid request body"}) return } if request.Scope != "internal" && request.Scope != "external" { writeJSON(w, http.StatusBadRequest, map[string]string{"error": "scope must be internal or external"}) return } if strings.TrimSpace(request.SnapshotID) == "" { writeJSON(w, http.StatusBadRequest, map[string]string{"error": "snapshot_id required"}) return } if body, held := instanceMaintenanceGuard(r.Context(), config, request.Scope); held { writeJSON(w, http.StatusConflict, body) return } digest, err := libraryRestoreParamsDigest(request.Scope, request.SnapshotID) if err != nil { writeJSON(w, http.StatusInternalServerError, map[string]string{"error": "cannot digest params"}) return } // 双闸与恢复演练证据同样适用于恢复:恢复也是写生产。 reasons := make([]string, 0, 4) if !config.ProductionReleaseRequestsEnabled { reasons = append(reasons, "生产写入闸口关闭") } if !config.ProductionBackupRestoreVerified { reasons = append(reasons, "没有已核验的生产恢复演练证据") } authorizationOK := false if storeWithAuth, ok := config.Store.(interface { ActiveLibraryAuthorization(ctx context.Context, scope, snapshotID, digest string, now time.Time) (governance.LibraryReleaseAuthorization, bool, error) }); ok { _, found, authErr := storeWithAuth.ActiveLibraryAuthorization(r.Context(), request.Scope, request.SnapshotID, digest, time.Now().UTC()) authorizationOK = authErr == nil && found } if !authorizationOK { reasons = append(reasons, "没有针对该快照的恢复授权") } if len(reasons) > 0 { writeJSON(w, http.StatusServiceUnavailable, map[string]any{"code": "library_restore_blocked", "reasons": reasons}) return } now := time.Now().UTC() jobID := "librestore-" + request.Scope + "-" + now.Format("20060102T150405.000000") job, reused, err := store.CreateLibraryJob(r.Context(), governance.LibraryJob{ ID: jobID, Scope: request.Scope, Kind: governance.LibraryJobRestore, IdempotencyKey: "library.restore:" + request.Scope + ":" + request.SnapshotID, RequestDigest: digest, TargetSnapshotID: request.SnapshotID, Actor: identity.Username, }, governance.SyncStages(), now) if errors.Is(err, governance.ErrLibraryJobActive) { writeJSON(w, http.StatusConflict, map[string]any{"code": "library_restore_active", "error": "该目标已有在飞作业"}) return } if errors.Is(err, governance.ErrLibraryJobIdempotencyConflict) { writeJSON(w, http.StatusConflict, map[string]any{"code": "library_restore_conflict", "error": "同一恢复请求参数不一致"}) return } if err != nil { writeJSON(w, http.StatusInternalServerError, map[string]string{"error": "cannot create restore job"}) return } if reused && (job.State == governance.LibraryJobFailed || job.State == governance.LibraryJobCancelled) { job, reused, err = store.CreateLibraryJob(r.Context(), governance.LibraryJob{ ID: jobID + "-retry", Scope: request.Scope, Kind: governance.LibraryJobRestore, IdempotencyKey: "library.restore:" + request.Scope + ":" + request.SnapshotID + ":retry:" + now.Format("20060102T150405.000000000"), RequestDigest: digest, TargetSnapshotID: request.SnapshotID, Actor: identity.Username, }, governance.SyncStages(), now) if err != nil || reused { writeJSON(w, http.StatusConflict, map[string]any{"code": "library_restore_retry_conflict", "error": "无法为重试创建新作业"}) return } } if reused { writeJSON(w, http.StatusAccepted, map[string]any{"job": job, "reused": true}) return } payload, err := json.Marshal(map[string]string{ "library_job_id": job.ID, "scope": request.Scope, "target_snapshot_id": request.SnapshotID, }) if err != nil { writeJSON(w, http.StatusInternalServerError, map[string]string{"error": "cannot build restore payload"}) return } queueJob, err := config.Store.Enqueue(r.Context(), jobs.EnqueueRequest{ ID: governance.QueueJobIDFor(job.ID), Type: libraryRestoreJobType, Payload: payload, AvailableAt: now, MaxAttempts: 1, IdempotencyKey: "library.restore.queue:" + job.ID, }, now) if err != nil || queueJob == nil { _ = store.CancelLibraryJob(r.Context(), job.ID, "队列投递失败,未开始恢复", now) writeJSON(w, http.StatusInternalServerError, map[string]string{"error": "cannot enqueue restore job"}) return } if attachErr := store.AttachLibraryQueueJob(r.Context(), job.ID, queueJob.ID, now); attachErr != nil { writeJSON(w, http.StatusInternalServerError, map[string]string{"error": "cannot attach restore job"}) return } writeJSON(w, http.StatusAccepted, map[string]any{"job": job, "reused": false}) } // 崩溃续作:失败作业可以按原参数重新排队,生成新作业(旧作业保留为审计记录)。 func libraryJobResume(w http.ResponseWriter, r *http.Request, config Config) { identity, ok := sessionFrom(r.Context()) // 续作会重新写生产,权限与恢复一致:只有管理员能触发。 if !ok || roleRank(identity.Role) < roleRank(roleAdmin) { writeJSON(w, http.StatusForbidden, map[string]any{"error": "续作需要管理员权限"}) return } store, ok := libraryStore(config) if !ok { writeJSON(w, http.StatusServiceUnavailable, map[string]any{"error": "作业存储不可用"}) return } sourceID := strings.TrimSpace(r.PathValue("id")) source, err := store.LibraryJob(r.Context(), sourceID) if err != nil { writeJSON(w, http.StatusNotFound, map[string]any{"error": "作业不存在"}) return } if source.State != governance.LibraryJobFailed && source.State != governance.LibraryJobCancelled { writeJSON(w, http.StatusConflict, map[string]any{"error": "只有失败或已取消的作业可以续作"}) return } // 与同步/恢复一致:闸口、恢复演练证据与授权都要在 API 侧 fail-closed, // 不能让前台接受一个 worker 必然拒绝的请求。 reasons := make([]string, 0, 3) if !config.ProductionReleaseRequestsEnabled { reasons = append(reasons, "生产写入闸口关闭") } if !config.ProductionBackupRestoreVerified { reasons = append(reasons, "没有已核验的生产恢复演练证据") } var digest string switch source.Kind { case governance.LibraryJobSync: digest, _ = librarySyncParamsDigest(source.Scope, source.TargetSnapshotID) case governance.LibraryJobRestore: digest, _ = libraryRestoreParamsDigest(source.Scope, source.TargetSnapshotID) } authorizationOK := false if storeWithAuth, ok := config.Store.(interface { ActiveLibraryAuthorization(ctx context.Context, scope, snapshotID, digest string, now time.Time) (governance.LibraryReleaseAuthorization, bool, error) }); ok && digest != "" { _, found, authErr := storeWithAuth.ActiveLibraryAuthorization(r.Context(), source.Scope, source.TargetSnapshotID, digest, time.Now().UTC()) authorizationOK = authErr == nil && found } if !authorizationOK { reasons = append(reasons, "没有针对该快照的有效授权") } if len(reasons) > 0 { writeJSON(w, http.StatusServiceUnavailable, map[string]any{"code": "library_resume_blocked", "reasons": reasons}) return } now := time.Now().UTC() // 幂等键按源作业固定:重复点击返回同一个续作作业,而不是每次新建。 jobID := sourceID + "-resume" var kind governance.LibraryJobKind queueType := "" switch source.Kind { case governance.LibraryJobSync: kind, queueType = governance.LibraryJobSync, librarySyncJobType case governance.LibraryJobRestore: kind, queueType = governance.LibraryJobRestore, libraryRestoreJobType default: writeJSON(w, http.StatusConflict, map[string]any{"error": "封存作业由封存入口重发,不走续作"}) return } job, reused, err := store.CreateLibraryJob(r.Context(), governance.LibraryJob{ ID: jobID, Scope: source.Scope, Kind: kind, IdempotencyKey: "library.resume:" + sourceID, RequestDigest: digest, TargetSnapshotID: source.TargetSnapshotID, Actor: identity.Username, }, governance.SyncStages(), now) if errors.Is(err, governance.ErrLibraryJobActive) { writeJSON(w, http.StatusConflict, map[string]any{"error": "该目标已有在飞作业"}) return } if errors.Is(err, governance.ErrLibraryJobIdempotencyConflict) { writeJSON(w, http.StatusConflict, map[string]any{"error": "同一续作请求参数不一致"}) return } if err != nil { writeJSON(w, http.StatusInternalServerError, map[string]string{"error": "cannot create resumed job"}) return } if reused { writeJSON(w, http.StatusAccepted, map[string]any{"job": job, "resumed_from": sourceID, "reused": true}) return } // 写入已完成、只差登记版本的情况:续作只做幂等收尾,不重写生产。 body := map[string]any{"library_job_id": job.ID, "scope": source.Scope, "target_snapshot_id": source.TargetSnapshotID} if source.Stage == "finalize" && source.SnapshotID != "" { body["finalize_only"] = true } payload, _ := json.Marshal(body) queueJob, err := config.Store.Enqueue(r.Context(), jobs.EnqueueRequest{ ID: governance.QueueJobIDFor(job.ID), Type: queueType, Payload: payload, AvailableAt: now, MaxAttempts: 1, IdempotencyKey: "library.resume.queue:" + job.ID, }, now) if err != nil || queueJob == nil { _ = store.CancelLibraryJob(r.Context(), job.ID, "队列投递失败", now) writeJSON(w, http.StatusInternalServerError, map[string]string{"error": "cannot enqueue resumed job"}) return } // 附不上队列号就等于作业没人跑:必须显式失败并取消,不能静默返回 202。 if attachErr := store.AttachLibraryQueueJob(r.Context(), job.ID, queueJob.ID, now); attachErr != nil { _ = store.CancelLibraryJob(r.Context(), job.ID, "队列归属写入失败", now) writeJSON(w, http.StatusInternalServerError, map[string]string{"error": "cannot attach resumed job"}) return } writeJSON(w, http.StatusAccepted, map[string]any{"job": job, "resumed_from": sourceID, "reused": false}) }