@powersync/service-module-mongodb-storage 0.19.0 → 0.21.0

package/src/storage/implementation/v3/CompactionLease.ts

@@ -0,0 +1,152 @@
+import { mongo } from '@powersync/lib-service-mongodb';
+import { InternalOpId } from '@powersync/service-core';
+import { CompactionLeaseLostError } from '../MongoCompactor.js';
+import { BucketStateDocumentV3 } from './models.js';
+
+const LEASE_RENEW_INTERVAL_MS = 60 * 1000;
+
+/**
+ * Filters for buckets that have no lease or an expired lease.
+ */
+export const AVAILABLE_LEASE_EXPR = {
+  // $$NOW is evaluated by MongoDB, avoiding lease expiry races due
+  // to clocks on separate compact workers.
+  $expr: {
+    $or: [{ $eq: [{ $type: '$compact_lease' }, 'missing'] }, { $lte: ['$compact_lease.expires_at', '$$NOW'] }]
+  }
+};
+
+/**
+ * Owns one V3 bucket-compaction lease, including its server-time renewal and
+ * the owner-fenced operations which release it.
+ *
+ * This is intended as a way to reduce redundant work if multiple jobs attempt to
+ * compact the same bucket concurrently, but it's not an absolute safety guarantee.
+ * The individual operations must still be designed to be safe with concurrent compacting.
+ */
+export class CompactionLease implements AsyncDisposable {
+  readonly startedAt: Date;
+  readonly lastOp: InternalOpId;
+
+  private timer: NodeJS.Timeout | undefined;
+  private renewalInFlight = false;
+  private renewalError: unknown;
+  private finalizing = false;
+  private finished = false;
+
+  private constructor(
+    private readonly collection: mongo.Collection<BucketStateDocumentV3>,
+    readonly state: BucketStateDocumentV3,
+    readonly id: mongo.ObjectId,
+    private readonly durationMs: number
+  ) {
+    this.startedAt = new Date(state.compact_lease!.expires_at.getTime() - durationMs);
+    this.lastOp = state.last_op;
+  }
+
+  static async claim(
+    collection: mongo.Collection<BucketStateDocumentV3>,
+    filter: mongo.Filter<BucketStateDocumentV3>,
+    sort: mongo.Sort | undefined,
+    durationMs: number
+  ): Promise<CompactionLease | null> {
+    const id = new mongo.ObjectId();
+    const state = await collection.findOneAndUpdate(
+      {
+        $and: [filter, AVAILABLE_LEASE_EXPR]
+      },
+      [
+        {
+          $set: {
+            compact_lease: {
+              id,
+              expires_at: { $dateAdd: { startDate: '$$NOW', unit: 'millisecond', amount: durationMs } }
+            }
+          }
+        }
+      ],
+      { sort, returnDocument: 'after' }
+    );
+    return state == null ? null : new CompactionLease(collection, state, id, durationMs);
+  }
+
+  startRenewal() {
+    const interval = Math.max(1, Math.min(LEASE_RENEW_INTERVAL_MS, Math.floor(this.durationMs / 2)));
+    this.timer = setInterval(() => {
+      if (this.renewalInFlight) {
+        return;
+      }
+      this.renewalInFlight = true;
+      void this.renew()
+        .catch((error) => {
+          this.renewalError = error;
+        })
+        .finally(() => {
+          this.renewalInFlight = false;
+        });
+    }, interval);
+    this.timer.unref();
+  }
+
+  async throwIfLost() {
+    if (this.renewalError != null) {
+      throw this.renewalError;
+    }
+  }
+
+  async reschedule(nextCompactCheck: mongo.Document) {
+    await this.finish([{ $set: { next_compact_check: nextCompactCheck } }, { $unset: 'compact_lease' }]);
+  }
+
+  async finalize(update: mongo.Document) {
+    await this.finish([{ $set: update }, { $unset: 'compact_lease' }]);
+  }
+
+  async [Symbol.asyncDispose]() {
+    if (this.timer != null) {
+      clearInterval(this.timer);
+      this.timer = undefined;
+    }
+    if (this.finished) {
+      return;
+    }
+    this.finalizing = true;
+    await this.collection.updateOne(this.filter, { $unset: { compact_lease: '' } });
+    this.finished = true;
+  }
+
+  private get filter(): mongo.Filter<BucketStateDocumentV3> {
+    return { _id: this.state._id, 'compact_lease.id': this.id };
+  }
+
+  private async renew() {
+    const result = await this.collection.updateOne(this.filter, [
+      {
+        $set: {
+          'compact_lease.expires_at': {
+            $dateAdd: { startDate: '$$NOW', unit: 'millisecond', amount: this.durationMs }
+          }
+        }
+      }
+    ]);
+    // A finalization in progress owns the authoritative fenced result.
+    if (result.matchedCount != 1 && !this.finalizing) {
+      throw new CompactionLeaseLostError(`Lost compaction lease for bucket ${this.state._id.b}`);
+    }
+    // A successful renewal confirms that a transient failure has passed. A
+    // lease-loss error remains sticky: it means another worker may own it.
+    if (!(this.renewalError instanceof CompactionLeaseLostError)) {
+      this.renewalError = undefined;
+    }
+  }
+
+  private async finish(update: mongo.Document[]) {
+    await this.throwIfLost();
+    this.finalizing = true;
+    const result = await this.collection.updateOne(this.filter, update);
+    if (result.matchedCount != 1) {
+      throw new CompactionLeaseLostError(`Lost compaction lease for bucket ${this.state._id.b}`);
+    }
+    this.finished = true;
+  }
+}

package/src/storage/implementation/v3/MongoBucketBatchV3.ts

@@ -1,21 +1,21 @@
 import * as lib_mongo from '@powersync/lib-service-mongodb';
-import { ReplicationAssertionError } from '@powersync/lib-services-framework';
-import { storage } from '@powersync/service-core';
+import { ReplicationAssertionError, ServiceAssertionError } from '@powersync/lib-services-framework';
+import { InternalOpId, storage } from '@powersync/service-core';
+import { EventDefinitionId } from '@powersync/service-sync-rules';
 import * as bson from 'bson';
 import { mongoTableId } from '../../../utils/util.js';
 import { canCheckpointState } from '../CheckpointState.js';
 import { MongoBucketBatch, MongoBucketBatchOptions } from '../MongoBucketBatch.js';
 import { MongoParsedSyncConfigSet } from '../MongoParsedSyncConfigSet.js';
-import { syncRuleStateUpdatePipeline } from '../SyncRuleStateUpdate.js';
+import { stopReplicationStreamPipeline } from '../SyncRuleStateUpdate.js';
 import { PersistedBatch } from '../common/PersistedBatch.js';
 import { SourceRecordStore } from '../common/SourceRecordStore.js';
 import { PersistedBatchV3 } from './PersistedBatchV3.js';
 import { SourceRecordStoreV3 } from './SourceRecordStoreV3.js';
 import { VersionedPowerSyncMongoV3 } from './VersionedPowerSyncMongoV3.js';
-import { ReplicationStreamDocumentV3, SourceTableDocumentV3 } from './models.js';
+import { CustomCheckpointRequestDocumentV3, ReplicationStreamDocumentV3, SourceTableDocumentV3 } from './models.js';
 import {
   createNewSourceTable,
-  designateEventCarrier,
   overlappingSourceTableFilter,
   planSourceTableReconciliation,
   sourceTableDesiredResolution,
@@ -27,6 +27,10 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
   declare readonly db: VersionedPowerSyncMongoV3;
 
   private readonly store: SourceRecordStore;
+  /** Avoid repeating idempotent createIndexes calls for every checkpoint flush. */
+  private readonly initializedCustomCheckpointEventIds = new Set<EventDefinitionId>();
+  // Used only to validate supplied checkpoint ids; it does not initialize collections.
+  private readonly knownEventIds: ReadonlySet<EventDefinitionId>;
   private readonly syncConfigIds: bson.ObjectId[];
   private needsActivationV3 = true;
   private lastWaitingLogThrottledV3 = 0;
@@ -38,6 +42,7 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
       throw new ReplicationAssertionError('Missing sync config id for v3 batch');
     }
     this.syncConfigIds = syncConfigIds;
+    this.knownEventIds = new Set(this.mapping.allEventDefinitionIds());
     this.store = new SourceRecordStoreV3(this.db, this.replicationStreamId, this.mapping);
   }
 
@@ -45,7 +50,9 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
     return new PersistedBatchV3(this.db, this.replicationStreamId, this.mapping, writtenSize, {
       logger: this.logger,
       objectStorage: this.options.objectStorage,
-      inlineThresholdBytes: this.options.inlineThresholdBytes
+      inlineThresholdBytes: this.options.inlineThresholdBytes,
+      signal: this.options.signal,
+      objectStorageUsageWriterId: this.objectStorageUsageWriterId
     });
   }
 
@@ -90,7 +97,8 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
         this.syncConfigIds.map((id) => id.toHexString()),
         table.ref,
         bucketDataSourceIds,
-        parameterLookupSourceIds
+        parameterLookupSourceIds,
+        [...table.eventDefinitionIds!]
       )
     );
   }
@@ -132,7 +140,7 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
         .sourceRecords(this.replicationStreamId, mongoTableId(table.id))
         .drop()
         .catch((error) => {
-          if (lib_mongo.isMongoServerError(error) && error.codeName === 'NamespaceNotFound') {
+          if (lib_mongo.isMongoNamespaceNotFoundError(error)) {
             return;
           }
           throw error;
@@ -146,25 +154,19 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
     const parsedOverride = options.parsedSyncConfig as MongoParsedSyncConfigSet | undefined;
     const syncConfig = parsedOverride?.hydratedSyncConfig ?? this.sync_rules;
     const mapping = parsedOverride?.mapping ?? this.mapping;
+    const eventById = (parsedOverride ?? this.options.parsedSyncConfig).eventById;
 
     const { connection_id, source } = options;
-    const context: SourceTableReconciliationContext = {
-      connectionId: connection_id,
-      connectionTag: source.connectionTag,
-      identity: {
-        schema: source.schema,
-        name: source.name,
-        objectId: source.objectId,
-        replicaIdColumns: source.replicaIdColumns.map((column) => ({
-          name: column.name,
-          type: column.type,
-          type_oid: column.typeId
-        }))
-      },
-      storeCurrentData: source.sendsCompleteRows !== true,
-      syncConfig,
-      mapping,
-      desired: sourceTableDesiredResolution(syncConfig, source, mapping)
+    const reconcile = options.reconcileSourceTables ?? storage.defaultSourceTableReconciler;
+    const identity = {
+      schema: source.schema,
+      name: source.name,
+      objectId: source.objectId,
+      replicaIdColumns: source.replicaIdColumns.map((column) => ({
+        name: column.name,
+        type: column.type,
+        type_oid: column.typeId
+      }))
     };
 
     let result: storage.ResolveTablesResult | null = null;
@@ -174,14 +176,36 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
     await session.withTransaction(async () => {
       const col = this.db.sourceTables(this.replicationStreamId);
 
-      // Fetch every persisted source-table doc that can overlap this physical table.
-      // Exact-identity docs are candidates for reuse; non-exact overlaps are possible drops.
+      // Find records that overlap by name or relation id.
       const candidateDocs = await col
-        .find(overlappingSourceTableFilter(connection_id, context.identity), { session })
+        .find(overlappingSourceTableFilter(connection_id, identity), { session })
         .toArray();
 
-      // Pure planning: which docs to retain (and narrow), whether a new doc is needed for
-      // uncovered memberships, and which docs conflict with the current identity.
+      const candidateTables = candidateDocs.map((doc) =>
+        sourceTableFromDocument(doc, source.connectionTag, syncConfig, mapping, eventById)
+      );
+      const candidates = candidateTables.map((table) => table.clone());
+      const resolution = await reconcile({ source, candidates });
+      storage.validateSourceTableCandidateResolution(candidates, resolution);
+
+      // Persist metadata from the reconciler without mutating the queried documents.
+      for (const { id, sourceMetadata } of storage.diffSourceTableUpdates(candidateTables, resolution)) {
+        await col.updateOne({ _id: mongoTableId(id) }, { $set: { source_metadata: sourceMetadata } }, { session });
+      }
+
+      const context: SourceTableReconciliationContext = {
+        connectionId: connection_id,
+        connectionTag: source.connectionTag,
+        identity,
+        storeCurrentData: source.sendsCompleteRows !== true,
+        syncConfig,
+        mapping,
+        desired: sourceTableDesiredResolution(syncConfig, source, mapping, eventById),
+        sourceCompatibleTables: resolution.compatibleTables,
+        newTableSourceMetadata: resolution.newTableValues.sourceMetadata
+      };
+
+      // Plan record reuse, membership changes, creation, and removal.
       const plan = planSourceTableReconciliation(candidateDocs, context);
 
       // Persist narrowing for incomplete snapshots only. Snapshot-complete docs keep stale
@@ -194,7 +218,8 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
           {
             $set: {
               bucket_data_source_ids: update.memberships.bucketDataSourceIds,
-              parameter_lookup_source_ids: update.memberships.parameterLookupSourceIds
+              parameter_lookup_source_ids: update.memberships.parameterLookupSourceIds,
+              event_definition_ids: update.memberships.eventDefinitionIds
             }
           },
           { session }
@@ -212,12 +237,11 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
         plan.tables.push(table);
       }
 
-      // If memberships are split across multiple source tables, only one may fire events.
-      designateEventCarrier(plan.tables, context.desired.triggersEvent);
-
       result = {
         tables: plan.tables,
-        dropTables: plan.dropDocs.map((doc) => sourceTableFromDocument(doc, context.connectionTag, syncConfig, mapping))
+        dropTables: plan.dropDocs.map((doc) =>
+          sourceTableFromDocument(doc, context.connectionTag, syncConfig, mapping, eventById)
+        )
       };
     });
 
@@ -232,12 +256,13 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
       return null;
     }
 
-    const refreshed = sourceTableFromDocument(doc, table.ref.connectionTag, this.sync_rules, this.mapping);
-    // The event-carrier designation is decided per resolveTables result and not persisted -
-    // preserve the caller's designation instead of recomputing it from the ref, so that
-    // refreshing a non-carrier table does not make it fire events.
-    refreshed.syncEvent = table.syncEvent;
-    return refreshed;
+    return sourceTableFromDocument(
+      doc,
+      table.ref.connectionTag,
+      this.sync_rules,
+      this.mapping,
+      this.options.parsedSyncConfig.eventById
+    );
   }
 
   async commit(lsn: string, options?: storage.BucketBatchCommitOptions): Promise<storage.CheckpointResult> {
@@ -504,7 +529,7 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
             _id: { $ne: this.replicationStreamId },
             state: { $in: [storage.SyncRuleState.ACTIVE, storage.SyncRuleState.ERRORED] }
           },
-          syncRuleStateUpdatePipeline(storage.SyncRuleState.STOP),
+          stopReplicationStreamPipeline(),
           { session }
         );
         activated = true;
@@ -668,4 +693,96 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
       return copy;
     });
   }
+
+  protected override async batchCreateCustomWriteCheckpoints(
+    session: lib_mongo.mongo.ClientSession,
+    opId: InternalOpId
+  ): Promise<void> {
+    if (this.write_checkpoint_batch.length == 0) {
+      return;
+    }
+
+    const checkpointsByEvent = Map.groupBy(this.write_checkpoint_batch, (checkpoint) =>
+      this.validateCustomCheckpointEventId(checkpoint.event_id)
+    );
+
+    for (const [eventId, checkpoints] of checkpointsByEvent) {
+      await this.batchCreateEventCustomWriteCheckpoints(session, opId, eventId, checkpoints);
+    }
+  }
+
+  protected override async prepareCustomWriteCheckpoints(): Promise<void> {
+    // Most events never produce custom checkpoints, so create collections lazily
+    // only for the event ids present in this batch. This hook runs before the
+    // replication transaction because MongoDB cannot create indexes within it.
+    const eventIds = new Set(
+      this.write_checkpoint_batch.map((checkpoint) => this.validateCustomCheckpointEventId(checkpoint.event_id))
+    );
+    for (const eventId of eventIds) {
+      if (this.initializedCustomCheckpointEventIds.has(eventId)) {
+        continue;
+      }
+      await this.db.initializeCustomCheckpointRequestsCollection({
+        replicationStreamId: this.replicationStreamId,
+        eventId
+      });
+      this.initializedCustomCheckpointEventIds.add(eventId);
+    }
+  }
+
+  private validateCustomCheckpointEventId(eventId: EventDefinitionId | undefined): EventDefinitionId {
+    if (eventId == null) {
+      throw new ServiceAssertionError('V3 custom checkpoints require an event definition id');
+    }
+
+    if (!this.knownEventIds.has(eventId)) {
+      throw new ServiceAssertionError(`Unknown custom checkpoint event definition ${eventId}`);
+    }
+    return eventId;
+  }
+
+  private async batchCreateEventCustomWriteCheckpoints(
+    session: lib_mongo.mongo.ClientSession,
+    opId: InternalOpId,
+    eventId: EventDefinitionId,
+    checkpoints: storage.CustomWriteCheckpointOptions[]
+  ): Promise<void> {
+    await this.db
+      .customCheckpointRequests({
+        eventId,
+        replicationStreamId: this.replicationStreamId
+      })
+      .bulkWrite(
+        checkpoints.map((checkpointOptions) => {
+          const set: Partial<CustomCheckpointRequestDocumentV3> = {
+            user_id: checkpointOptions.user_id,
+            checkpoint: checkpointOptions.checkpoint,
+            op_id: opId
+          };
+          if (checkpointOptions.checkpoint_requested_at != null) {
+            set.checkpoint_requested_at = checkpointOptions.checkpoint_requested_at;
+          }
+
+          return {
+            updateOne: {
+              filter: {
+                user_id: checkpointOptions.user_id
+              },
+              update: {
+                $set: set,
+                ...(checkpointOptions.checkpoint_requested_at == null
+                  ? {
+                      $unset: {
+                        checkpoint_requested_at: 1
+                      }
+                    }
+                  : {})
+              },
+              upsert: true
+            }
+          };
+        }),
+        { session }
+      );
+  }
 }