@friggframework/core 2.0.0-next.5 → 2.0.0-next.50

package/database/MONGODB_TRANSACTION_FIX.md

@@ -0,0 +1,198 @@
+# MongoDB Transaction Namespace Fix
+
+## Problem
+
+The encryption health check was failing with the following error:
+
+```
+Cannot create namespace frigg.Credential in multi-document transaction.
+Error code: 263
+```
+
+### Root Cause
+
+MongoDB does not allow creating collections (namespaces) inside multi-document transactions. When Prisma tries to create a document in a collection that doesn't exist yet, MongoDB needs to implicitly create the collection. If this happens inside a transaction context, MongoDB throws error code 263.
+
+### Technical Details
+
+- **MongoDB Constraint**: Collections must exist before being used in multi-document transactions
+- **Prisma Behavior**: Prisma may implicitly use transactions for certain operations
+- **Impact**: Health checks fail on fresh databases or when collections haven't been created yet
+
+## Solution
+
+**Implemented a comprehensive schema initialization system that ensures all collections exist at application startup.**
+
+### Architectural Approach
+
+Rather than checking before each individual database operation, we take a **systematic, fail-fast approach**:
+
+1. **Parse Prisma Schema**: Extract all collection names from the Prisma schema definition
+2. **Initialize at Startup**: Create all collections when the database connection is established
+3. **Fail Fast**: If there are database issues, the application fails immediately at startup rather than during runtime operations
+4. **Idempotent**: Safe to run multiple times - only creates collections that don't exist
+
+This follows the **"fail fast"** principle and ensures consistent state across all application instances.
+
+### Changes Made
+
+1. **Created MongoDB Schema Initialization** (`packages/core/database/utils/mongodb-schema-init.js`)
+   - `initializeMongoDBSchema()` - Ensures all Prisma collections exist at startup
+   - `getPrismaCollections()` - Returns list of all Prisma collection names
+   - `PRISMA_COLLECTIONS` - Constant array of all 13 Prisma collections
+   - Only runs for MongoDB (skips PostgreSQL)
+   - Fails fast if database not connected
+
+2. **Created MongoDB Collection Utilities** (`packages/core/database/utils/mongodb-collection-utils.js`)
+   - `ensureCollectionExists(collectionName)` - Ensures a single collection exists
+   - `ensureCollectionsExist(collectionNames)` - Batch creates multiple collections
+   - `collectionExists(collectionName)` - Checks if a collection exists
+   - Handles race conditions gracefully (NamespaceExists errors)
+
+3. **Integrated into Database Connection** (`packages/core/database/prisma.js`)
+   - Modified `connectPrisma()` to call `initializeMongoDBSchema()` after connection
+   - Ensures all collections exist before application handles requests
+
+4. **Updated Health Check Repository** (`packages/core/database/repositories/health-check-repository-mongodb.js`)
+   - Removed per-operation collection existence checks
+   - Added documentation noting schema is initialized at startup
+
+5. **Added Comprehensive Tests**
+   - `mongodb-schema-init.test.js` - Tests schema initialization system
+   - `mongodb-collection-utils.test.js` - Tests collection utility functions
+   - Tests error handling, race conditions, and edge cases
+
+### Implementation Flow
+
+```javascript
+// 1. Application startup - connect to database
+await connectPrisma();
+  └─> await initializeMongoDBSchema();
+       └─> await ensureCollectionsExist([
+            'User', 'Token', 'Credential', 'Entity',
+            'Integration', 'IntegrationMapping', 'Process',
+            'Sync', 'DataIdentifier', 'Association',
+            'AssociationObject', 'State', 'WebsocketConnection'
+           ]);
+
+// 2. Now all collections exist - safe to handle requests
+// No per-operation checks needed!
+await prisma.credential.create({ data: {...} }); // Works without namespace error
+```
+
+## Best Practices Followed
+
+1. **Domain-Driven Design**: Created reusable utility module for MongoDB-specific concerns
+2. **Hexagonal Architecture**: Infrastructure concerns (schema initialization) handled in infrastructure layer
+3. **Test-Driven Development**: Added comprehensive tests for all utility functions
+4. **Fail Fast Principle**: Database issues discovered at startup, not during runtime
+5. **Idempotency**: Safe to run multiple times across multiple instances
+6. **Error Handling**: Graceful degradation on race conditions and errors
+7. **Documentation**: Inline comments, JSDoc, and comprehensive documentation
+
+## Benefits
+
+### Immediate Benefits
+- ✅ Fixes encryption health check failures on fresh databases
+- ✅ Prevents transaction namespace errors across **all** Prisma operations
+- ✅ No per-operation overhead - collections created once at startup
+- ✅ Fail fast - database issues discovered immediately at startup
+- ✅ Idempotent - safe to run multiple times and across multiple instances
+
+### Architectural Benefits
+- ✅ **Clean separation of concerns**: Schema initialization is infrastructure concern, handled at startup
+- ✅ **Follows DDD/Hexagonal Architecture**: Infrastructure layer handles database setup, repositories focus on business operations
+- ✅ **Consistent across all environments**: Dev, test, staging, production all follow same pattern
+- ✅ **No repository-level checks needed**: All repositories benefit automatically
+- ✅ **Well-tested and documented**: Comprehensive test coverage and documentation
+
+### Operational Benefits
+- ✅ **Predictable startup**: Clear logging of schema initialization
+- ✅ **Zero runtime overhead**: Collections created once, not on every operation
+- ✅ **Production-ready**: Handles race conditions, errors, and edge cases gracefully
+
+## Design Decisions
+
+### Why Initialize at Startup?
+
+We considered two approaches:
+
+**❌ Per-Operation Checks (Initial approach)**
+```javascript
+async createCredential(data) {
+    await ensureCollectionExists('Credential'); // Check every time
+    return await prisma.credential.create({ data });
+}
+```
+- Pros: Guarantees collection exists before each operation
+- Cons: Runtime overhead, repeated checks, scattered logic
+
+**✅ Startup Initialization (Final approach)**
+```javascript
+// Once at startup
+await connectPrisma(); // Initializes all collections
+
+// All operations just work
+async createCredential(data) {
+    return await prisma.credential.create({ data }); // No checks needed
+}
+```
+- Pros: Zero runtime overhead, centralized logic, fail fast, consistent
+- Cons: Requires database connection at startup (already required)
+
+### Benefits of Startup Approach
+
+1. **Performance**: Collections created once vs. checking before every operation
+2. **Simplicity**: No conditional logic in repositories
+3. **Reliability**: Fail fast at startup if database has issues
+4. **Maintainability**: Single source of truth for schema initialization
+5. **DDD Alignment**: Infrastructure concerns handled in infrastructure layer
+
+## Logging Output
+
+When the application starts, you'll see clear logging:
+
+```
+Initializing MongoDB schema - ensuring all collections exist...
+Created MongoDB collection: Credential
+MongoDB schema initialization complete - 13 collections verified (45ms)
+```
+
+On subsequent startups (collections already exist):
+```
+Initializing MongoDB schema - ensuring all collections exist...
+MongoDB schema initialization complete - 13 collections verified (12ms)
+```
+
+## References
+
+- [Prisma Issue #8305](https://github.com/prisma/prisma/issues/8305) - MongoDB "Cannot create namespace" error
+- [Mongoose Issue #6699](https://github.com/Automattic/mongoose/issues/6699) - Similar issue in Mongoose
+- [MongoDB Transactions Documentation](https://www.mongodb.com/docs/manual/core/transactions/#transactions-and-operations) - Operations allowed in transactions
+- [Prisma MongoDB Guide](https://www.prisma.io/docs/guides/database/mongodb) - Using Prisma with MongoDB
+
+## Future Considerations
+
+### Automatic Schema Sync
+Consider enhancing the system to:
+- Parse Prisma schema file dynamically to extract collection names
+- Auto-detect schema changes and create new collections
+- Provide CLI command for manual schema initialization
+
+### Migration Support
+For production deployments with existing data:
+- Document migration procedures for new collections
+- Consider pre-migration scripts for blue-green deployments
+- Add health check for schema initialization status
+
+### Multi-Database Support
+The system already handles:
+- ✅ MongoDB - Full schema initialization
+- ✅ PostgreSQL - Skips initialization (uses Prisma migrations)
+- Consider adding explicit migration support for DocumentDB-specific features
+
+### Index Creation
+Future enhancement could also create indexes at startup:
+- Parse Prisma schema for `@@index` directives
+- Create indexes if they don't exist
+- Provide index health checks

package/database/adapters/lambda-invoker.js

@@ -0,0 +1,97 @@
+/**
+ * Lambda Invoker Adapter
+ * Infrastructure layer - handles AWS Lambda function invocations
+ * 
+ * Part of Hexagonal Architecture:
+ * - Infrastructure Layer adapter for AWS SDK
+ * - Used by Domain Layer use cases
+ * - Isolates AWS-specific logic from business logic
+ */
+
+const { LambdaClient, InvokeCommand } = require('@aws-sdk/client-lambda');
+
+/**
+ * Custom error for Lambda invocation failures
+ * Provides structured error information for debugging
+ */
+class LambdaInvocationError extends Error {
+    constructor(message, functionName, statusCode) {
+        super(message);
+        this.name = 'LambdaInvocationError';
+        this.functionName = functionName;
+        this.statusCode = statusCode;
+    }
+}
+
+/**
+ * Adapter for invoking AWS Lambda functions
+ * 
+ * Infrastructure layer - handles AWS SDK communication
+ * Converts AWS SDK responses to domain-friendly formats
+ */
+class LambdaInvoker {
+    /**
+     * @param {LambdaClient} lambdaClient - AWS Lambda client (injected for testability)
+     */
+    constructor(lambdaClient = new LambdaClient({})) {
+        this.client = lambdaClient;
+    }
+
+    /**
+     * Invoke Lambda function synchronously
+     * 
+     * @param {string} functionName - Lambda function name or ARN
+     * @param {Object} payload - Event payload to send to Lambda
+     * @returns {Promise<Object>} Parsed response body
+     * @throws {LambdaInvocationError} If Lambda returns error status
+     * @throws {Error} If AWS SDK call fails
+     */
+    async invoke(functionName, payload) {
+        try {
+            const command = new InvokeCommand({
+                FunctionName: functionName,
+                InvocationType: 'RequestResponse', // Synchronous
+                Payload: JSON.stringify(payload),
+            });
+
+            const response = await this.client.send(command);
+
+            // Parse response payload
+            let result;
+            try {
+                result = JSON.parse(Buffer.from(response.Payload).toString());
+            } catch (parseError) {
+                throw new LambdaInvocationError(
+                    `Failed to parse Lambda response: ${parseError.message}`,
+                    functionName,
+                    null
+                );
+            }
+
+            // Check status code
+            if (result.statusCode === 200) {
+                return result.body;
+            }
+
+            // Lambda returned error status
+            const errorMessage = result.body?.error || 'Lambda invocation failed';
+            throw new LambdaInvocationError(
+                `Lambda ${functionName} returned error: ${errorMessage}`,
+                functionName,
+                result.statusCode
+            );
+        } catch (error) {
+            // Re-throw LambdaInvocationError as-is
+            if (error instanceof LambdaInvocationError) {
+                throw error;
+            }
+
+            // Wrap AWS SDK errors
+            throw new Error(`Failed to invoke Lambda ${functionName}: ${error.message}`);
+        }
+    }
+}
+
+module.exports = { LambdaInvoker, LambdaInvocationError };
+
+

package/database/config.js

@@ -0,0 +1,154 @@
+/**
+ * Database Configuration
+ * Manages configuration for Prisma ORM operations
+ */
+
+/**
+ * Determines database type from environment or app definition
+ * 
+ * Detection order:
+ * 1. DB_TYPE environment variable (set for migration handlers)
+ * 2. App definition (backend/index.js Definition.database configuration)
+ *
+ * @returns {'mongodb'|'postgresql'} Database type
+ * @throws {Error} If database type cannot be determined or app definition missing
+ */
+function getDatabaseType() {
+    // First, check DB_TYPE environment variable (migration handlers set this)
+    if (process.env.DB_TYPE) {
+        return process.env.DB_TYPE;
+    }
+
+    // Fallback: Load app definition
+    try {
+        const path = require('node:path');
+        const fs = require('node:fs');
+        const { findNearestBackendPackageJson } = require('../utils');
+
+        let backendIndexPath;
+        let database;
+        const backendPackagePath = findNearestBackendPackageJson();
+
+        if (!backendPackagePath) {
+            throw new Error(
+                '[Frigg] Cannot find backend package.json. ' +
+                'Ensure backend/package.json exists in your project.'
+            );
+        }
+
+        const backendDir = path.dirname(backendPackagePath);
+        backendIndexPath = path.join(backendDir, 'index.js');
+
+        if (!fs.existsSync(backendIndexPath)) {
+            throw new Error(
+                `[Frigg] Backend index.js not found at ${backendIndexPath}. ` +
+                'Ensure backend/index.js exists with a Definition export.'
+            );
+        }
+
+        let backendModule;
+        try {
+            backendModule = require(backendIndexPath);
+        } catch (requireError) {
+            // Extract the actual file with the error from the stack trace
+            // Skip internal Node.js files (node:internal/*) and find first user file
+            let errorFile = 'unknown file';
+            const stackLines = requireError.stack?.split('\n') || [];
+
+            for (const line of stackLines) {
+                // Match file paths in stack trace, excluding node:internal
+                const match = line.match(/\(([^)]+\.js):\d+:\d+\)/) || line.match(/at ([^(]+\.js):\d+:\d+/);
+                if (match && match[1] && !match[1].includes('node:internal')) {
+                    errorFile = match[1];
+                    break;
+                }
+            }
+
+            // Provide better error context for syntax/runtime errors
+            throw new Error(
+                `[Frigg] Failed to load app definition from ${backendIndexPath}\n` +
+                `Error: ${requireError.message}\n` +
+                `File with error: ${errorFile}\n` +
+                `\nFull stack trace:\n${requireError.stack}\n\n` +
+                'This error occurred while loading your app definition or its dependencies. ' +
+                'Check the file listed above for syntax errors (trailing commas, missing brackets, etc.)'
+            );
+        }
+
+        database = backendModule?.Definition?.database;
+
+        if (!database) {
+            throw new Error(
+                '[Frigg] App definition missing database configuration. ' +
+                `Add database: { postgres: { enable: true } } (or mongoDB/documentDB) to ${backendIndexPath}`
+            );
+        }
+
+        // Determine database type from enabled database
+        // Priority order: postgres > mongoDB > documentDB
+        if (database.postgres?.enable === true) {
+            return 'postgresql';
+        }
+        if (database.mongoDB?.enable === true) {
+            return 'mongodb';
+        }
+        if (database.documentDB?.enable === true) {
+            return 'mongodb'; // DocumentDB is MongoDB-compatible
+        }
+
+        throw new Error(
+            '[Frigg] No database enabled in app definition. ' +
+            'Set one of: database.postgres.enable, database.mongoDB.enable, or database.documentDB.enable to true'
+        );
+    } catch (error) {
+        // Re-throw with context if it's our error
+        if (error.message.includes('[Frigg]')) {
+            throw error;
+        }
+        // Wrap unexpected errors
+        throw new Error(
+            `[Frigg] Failed to determine database type: ${error.message}`
+        );
+    }
+}
+
+/**
+ * Cached database type (lazy evaluation)
+ * @type {'mongodb'|'postgresql'|null}
+ */
+let cachedDbType = null;
+
+/**
+ * Enable Prisma debug logging
+ * Set PRISMA_LOG_LEVEL to comma-separated list: query,info,warn,error
+ * @type {string}
+ */
+const PRISMA_LOG_LEVEL = process.env.PRISMA_LOG_LEVEL || 'error,warn';
+
+/**
+ * Enable Prisma query logging for performance monitoring
+ * @type {boolean}
+ */
+const PRISMA_QUERY_LOGGING = process.env.PRISMA_QUERY_LOGGING === 'true';
+
+module.exports = {
+    getDatabaseType, // Export for testing and direct use
+    PRISMA_LOG_LEVEL,
+    PRISMA_QUERY_LOGGING,
+};
+
+/**
+ * Lazy-evaluated database type determined from app definition
+ * Only evaluates when accessed, preventing module load failures in test environments
+ * @type {'mongodb'|'postgresql'}
+ */
+Object.defineProperty(module.exports, 'DB_TYPE', {
+    get() {
+        if (cachedDbType === null) {
+            cachedDbType = getDatabaseType();
+        }
+        return cachedDbType;
+    },
+    enumerable: true,
+    configurable: true
+});