@powersync/service-module-mongodb-storage 0.20.0 → 0.22.0

package/src/storage/implementation/v1/PersistedBatchV1.ts

@@ -6,6 +6,7 @@ import * as bson from 'bson';
 import { BucketDataSource, BucketDefinitionId } from '@powersync/service-sync-rules';
 import { mongoTableId } from '../../../utils/util.js';
 import { EMPTY_DATA } from '../MongoBucketBatchShared.js';
+import { MongoWriteBatch } from '../MongoWriteBatch.js';
 import {
   BucketStateUpdate,
   PersistedBatch,
@@ -25,7 +26,9 @@ import {
 export class PersistedBatchV1 extends PersistedBatch {
   declare protected readonly db: VersionedPowerSyncMongoV1;
 
-  currentData: mongo.AnyBulkWriteOperation<CurrentDataDocument>[] = [];
+  // Each upsert supplies the complete source-record state. Only its final
+  // state matters within a flush; bucket and parameter history is kept separately.
+  currentData = new Map<string, mongo.AnyBulkWriteOperation<CurrentDataDocument>>();
 
   protected checkDefinitionId(_definitionId: BucketDefinitionId | null): BucketDefinitionId {
     // V1 storage doesn't persist the id, and we don't use it.
@@ -94,7 +97,7 @@ export class PersistedBatchV1 extends PersistedBatch {
   }
 
   hardDeleteCurrentData(sourceTableId: bson.ObjectId, replicaId: storage.ReplicaId) {
-    this.currentData.push({
+    this.currentData.set(this.currentDataKey(sourceTableId, replicaId), {
       deleteOne: {
         filter: { _id: this.currentDataId(sourceTableId, replicaId) }
       }
@@ -124,7 +127,7 @@ export class PersistedBatchV1 extends PersistedBatch {
       return lookup.lookup;
     });
 
-    this.currentData.push({
+    this.currentData.set(this.currentDataKey(values.sourceTableId, values.replicaId), {
       updateOne: {
         filter: { _id: this.currentDataId(values.sourceTableId, values.replicaId) },
         update: {
@@ -141,57 +144,45 @@ export class PersistedBatchV1 extends PersistedBatch {
   }
 
   protected get currentDataCount() {
-    return this.currentData.length;
+    return this.currentData.size;
   }
 
-  protected async flushBucketData(session: mongo.ClientSession) {
-    await this.db.bucketDataV1.bulkWrite(
+  protected async queueBucketData(writes: MongoWriteBatch) {
+    writes.bulkWriteUnordered(
+      this.db.bucketDataV1,
       this.bucketData.map((document) => ({
         insertOne: {
           document: serializeBucketDataV1(document)
         }
-      })),
-      {
-        session,
-        ordered: false
-      }
+      }))
     );
   }
 
-  protected async flushBucketParameters(session: mongo.ClientSession) {
-    await this.db.parameterIndexV1.bulkWrite(
+  protected queueBucketParameters(writes: MongoWriteBatch): void {
+    writes.bulkWriteUnordered(
+      this.db.parameterIndexV1,
       this.bucketParameters.map((document) => ({
         insertOne: {
           document: taggedBucketParameterDocumentToV1(document)
         }
-      })),
-      {
-        session,
-        ordered: false
-      }
+      }))
     );
   }
 
-  protected async flushCurrentData(session: mongo.ClientSession) {
-    if (this.currentData.length == 0) {
+  protected queueCurrentData(writes: MongoWriteBatch): void {
+    if (this.currentData.size == 0) {
       return;
     }
 
-    await this.db.sourceRecordsV1.bulkWrite(this.currentData, {
-      session,
-      ordered: true
-    });
+    writes.bulkWriteUnordered(this.db.sourceRecordsV1, [...this.currentData.values()]);
   }
 
-  protected async flushBucketStates(session: mongo.ClientSession) {
-    await this.db.bucketStateV1.bulkWrite(this.getBucketStateUpdates(), {
-      session,
-      ordered: false
-    });
+  protected queueBucketStates(writes: MongoWriteBatch): void {
+    writes.bulkWriteUnordered(this.db.bucketStateV1, this.getBucketStateUpdates());
   }
 
   protected resetCurrentData() {
-    this.currentData = [];
+    this.currentData.clear();
   }
 
   private getBucketStateUpdates(): mongo.AnyBulkWriteOperation<BucketStateDocumentV1>[] {
@@ -219,6 +210,10 @@ export class PersistedBatchV1 extends PersistedBatch {
     });
   }
 
+  private currentDataKey(sourceTableId: bson.ObjectId, replicaId: storage.ReplicaId): string {
+    return Buffer.from(bson.serialize({ t: sourceTableId, k: replicaId })).toString('base64');
+  }
+
   private currentDataId(sourceTableId: bson.ObjectId, replicaId: storage.ReplicaId): SourceKey {
     return {
       g: this.group_id,

package/src/storage/implementation/v1/models.ts

@@ -48,6 +48,7 @@ export function serializeBucketDataV1(document: BucketDataDoc): BucketDataDocume
     op: document.op,
     source_table: document.source_table,
     source_key: document.source_key,
+    subkey: document.subkey,
     table: document.table,
     row_id: document.row_id,
     checksum: document.checksum,

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 +1,22 @@
 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 { MongoWriteBatch } from '../MongoWriteBatch.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 +28,10 @@ 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 +43,7 @@ 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 +51,9 @@ 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
     });
   }
 
@@ -53,14 +61,12 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
     return this.store;
   }
 
-  protected override async onReplicationTransactionFlush(
-    session: lib_mongo.mongo.ClientSession,
-    lastOp: bigint
-  ): Promise<void> {
+  protected override onReplicationTransactionFlush(writes: MongoWriteBatch, lastOp: bigint): void {
     // Durably advance the stream-level head of persisted ops within the flush transaction.
     // This ensures a checkpoint created later (even by an empty commit, or by a freshly-appended
     // config that replicates nothing) covers all ops persisted before a potential crash.
-    await this.db.sync_rules.updateOne(
+    writes.updateOne(
+      this.db.sync_rules,
       {
         _id: this.replicationStreamId
       },
@@ -68,8 +74,7 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
         $max: {
           last_persisted_op: lastOp
         }
-      },
-      { session }
+      }
     );
   }
 
@@ -90,7 +95,8 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
         this.syncConfigIds.map((id) => id.toHexString()),
         table.ref,
         bucketDataSourceIds,
-        parameterLookupSourceIds
+        parameterLookupSourceIds,
+        [...table.eventDefinitionIds!]
       )
     );
   }
@@ -132,7 +138,7 @@ 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;
@@ -146,6 +152,7 @@ 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 reconcile = options.reconcileSourceTables ?? storage.defaultSourceTableReconciler;
@@ -173,7 +180,7 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
         .toArray();
 
       const candidateTables = candidateDocs.map((doc) =>
-        sourceTableFromDocument(doc, source.connectionTag, syncConfig, mapping)
+        sourceTableFromDocument(doc, source.connectionTag, syncConfig, mapping, eventById)
       );
       const candidates = candidateTables.map((table) => table.clone());
       const resolution = await reconcile({ source, candidates });
@@ -191,7 +198,7 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
         storeCurrentData: source.sendsCompleteRows !== true,
         syncConfig,
         mapping,
-        desired: sourceTableDesiredResolution(syncConfig, source, mapping),
+        desired: sourceTableDesiredResolution(syncConfig, source, mapping, eventById),
         sourceCompatibleTables: resolution.compatibleTables,
         newTableSourceMetadata: resolution.newTableValues.sourceMetadata
       };
@@ -209,7 +216,8 @@ 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 }
@@ -227,12 +235,11 @@ 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)
+        )
       };
     });
 
@@ -247,12 +254,13 @@ 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> {
@@ -519,7 +527,7 @@ 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;
@@ -642,7 +650,9 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
     const syncConfigIds = this.relevantSyncConfigIdsForTables(tables);
 
     await this.withTransaction(async () => {
-      await this.db.sourceTables(this.replicationStreamId).updateMany(
+      const writes = this.db.createWriteBatch(session, { ordered: false });
+      writes.updateMany(
+        this.db.sourceTables(this.replicationStreamId),
         { _id: { $in: ids } },
         {
           $set: {
@@ -651,12 +661,12 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
           $unset: {
             snapshot_status: 1
           }
-        },
-        { session }
+        }
       );
 
       if (no_checkpoint_before_lsn != null && syncConfigIds.length > 0) {
-        await this.db.sync_rules.updateOne(
+        writes.updateOne(
+          this.db.sync_rules,
           {
             _id: this.replicationStreamId,
             'sync_configs._id': { $in: syncConfigIds }
@@ -670,12 +680,12 @@ export class MongoBucketBatchV3 extends MongoBucketBatch {
             }
           },
           {
-            session: this.session,
             // Only set for sync configs that use this table
             arrayFilters: [{ 'config._id': { $in: syncConfigIds } }]
           }
         );
       }
+      await writes.execute();
     });
     return tables.map((table) => {
       const copy = table.clone();
@@ -683,4 +693,98 @@ 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)
+    );
+
+    const writes = this.db.createWriteBatch(session, { ordered: false });
+    for (const [eventId, checkpoints] of checkpointsByEvent) {
+      this.batchCreateEventCustomWriteCheckpoints(writes, opId, eventId, checkpoints);
+    }
+    await writes.execute();
+  }
+
+  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 batchCreateEventCustomWriteCheckpoints(
+    writes: MongoWriteBatch,
+    opId: InternalOpId,
+    eventId: EventDefinitionId,
+    checkpoints: storage.CustomWriteCheckpointOptions[]
+  ): void {
+    // A repeated user within an event replaces the complete checkpoint state.
+    const uniqueCheckpoints = new Map(checkpoints.map((checkpoint) => [checkpoint.user_id, checkpoint]));
+    writes.bulkWriteUnordered(
+      this.db.customCheckpointRequests({
+        eventId,
+        replicationStreamId: this.replicationStreamId
+      }),
+      [...uniqueCheckpoints.values()].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
+          }
+        };
+      })
+    );
+  }
 }