11 KiB

Raw Blame History

Active Identity Implementation Overview

Author: Matthew Raymer Date: 2025-08-21T13:40Z Status: 🚧 IN PROGRESS - Implementation Complete, Testing Pending

Objective

Separate the activeDid field from the monolithic settings table into a dedicated active_identity table to achieve:

Data normalization and reduced cache drift
Simplified identity management with dedicated table
Zero breaking API surface for existing components
Phased migration with rollback capability

Result

This document provides a comprehensive overview of the implemented Active Identity table separation system, including architecture, migration strategy, and component integration.

Use/Run

The implementation is ready for testing. Components can immediately use the new façade methods while maintaining backward compatibility through dual-write triggers.

Context & Scope

Audience: Developers working with identity management and database migrations
In scope: Active DID management, database schema evolution, Vue component integration
Out of scope: Multi-profile support beyond basic scope framework, complex identity hierarchies

Artifacts & Links

Implementation: src/db/tables/activeIdentity.ts, src/utils/PlatformServiceMixin.ts
Migrations: src/db-sql/migration.ts (migrations 003 & 004)
Configuration: src/config/featureFlags.ts
Documentation: This document and progress tracking

Environment & Preconditions

Database: SQLite (Absurd-SQL for Web, Capacitor SQLite for Mobile)
Framework: Vue.js with PlatformServiceMixin
Migration System: Built-in migrationService.ts with automatic execution

Architecture / Process Overview

The Active Identity separation follows a phased migration pattern with dual-write triggers to ensure zero downtime and backward compatibility.

flowchart TD
    A[Legacy State] --> B[Phase A: Dual-Write]
    B --> C[Phase B: Component Cutover]
    C --> D[Phase C: Legacy Cleanup]

    A --> A1[settings.activeDid]
    B --> B1[active_identity table]
    B --> B2[Dual-write trigger]
    B --> B3[Fallback support]
    C --> C1[Components use façade]
    C --> C2[Legacy fallback disabled]
    D --> D1[Drop activeDid column]
    D --> D2[Remove triggers]

Interfaces & Contracts

Database Schema

Table	Purpose	Key Fields	Constraints
`active_identity`	Store active DID	`id`, `active_did`,	FK to accounts.did
		`updated_at`

Service Façade API

Method	Purpose	Parameters	Returns
`$getActiveDid()`	Retrieve active DID	None	`Promise<string \| null>`
`$setActiveDid(did)`	Set active DID	`did`	`Promise<void>`
`$switchActiveIdentity(did)`	Switch to different DID	`did`	`Promise<void>`
`$getActiveIdentityScopes()`	Get available scopes	None	`Promise<string[]>` (always returns `["default"]`)

Repro: End-to-End Procedure

1. Database Migration Execution

# Migrations run automatically on app startup
# Migration 003: Creates active_identity table
# Migration 004: Drops settings.activeDid column (Phase C)

2. Component Usage

// Before (legacy)
const activeDid = settings.activeDid || "";
await this.$saveSettings({ activeDid: newDid });

// After (new façade)
const activeDid = await this.$getActiveDid() || "";
await this.$setActiveDid(newDid);

3. Feature Flag Control

// Enable/disable migration phases
FLAGS.USE_ACTIVE_IDENTITY_ONLY = false;      // Allow legacy fallback
FLAGS.DROP_SETTINGS_ACTIVEDID = false;       // Keep legacy column
FLAGS.LOG_ACTIVE_ID_FALLBACK = true;         // Log fallback usage

What Works (Evidence)

✅ Migration Infrastructure: Migrations 003 and 004 integrated into migrationService.ts
✅ Table Creation: active_identity table schema with proper constraints and indexes
✅ Service Façade: PlatformServiceMixin extended with all required methods
✅ Feature Flags: Comprehensive flag system for controlling rollout phases
✅ Dual-Write Support: One-way trigger from settings.activeDid → active_identity.active_did
✅ Validation: DID existence validation before setting as active
✅ Error Handling: Comprehensive error handling with logging

What Doesn't (Evidence & Hypotheses)

❌ Component Migration: No components yet updated to use new façade methods
❌ Testing: No automated tests for new functionality
❌ Performance Validation: No benchmarks for read/write performance
❌ Cross-Platform Validation: Not tested on mobile platforms yet

Risks, Limits, Assumptions

Migration Risks

Data Loss: If migration fails mid-process, could lose active DID state
Rollback Complexity: Phase C (column drop) requires table rebuild, not easily reversible
Trigger Dependencies: Dual-write trigger could fail if active_identity table is corrupted

Performance Limits

Dual-Write Overhead: Each activeDid change triggers additional database operations
Fallback Queries: Legacy fallback requires additional database queries
Transaction Scope: Active DID changes wrapped in transactions for consistency

Security Boundaries

DID Validation: Only validates DID exists in accounts table, not ownership
Scope Isolation: No current scope separation enforcement beyond table constraints
Access Control: No row-level security on active_identity table

Next Steps

Owner	Task	Exit Criteria	Target Date (UTC)
Developer	Test migrations	Migrations execute without errors	2025-08-21
Developer	Update components	All components use new façade	2025-08-22
		methods
Developer	Performance testing	Read/write performance meets	2025-08-23
		requirements
Developer	Phase C activation	Feature flag enables column	2025-08-24
		removal

References

Competence Hooks

Why this works: Phased migration with dual-write triggers ensures zero downtime while maintaining data consistency through foreign key constraints and validation
Common pitfalls: Forgetting to update components before enabling USE_ACTIVE_IDENTITY_ONLY, not testing rollback scenarios, ignoring cross-platform compatibility
Next skill unlock: Implement automated component migration using codemods and ESLint rules
Teach-back: Explain how the dual-write trigger prevents data divergence during the transition phase

Collaboration Hooks

Reviewers: Database team for migration logic, Vue team for component integration, DevOps for deployment strategy
Sign-off checklist: Migrations tested in staging, components updated, performance validated, rollback plan documented

Assumptions & Limits

Single User Focus: Current implementation assumes single-user mode with 'default' scope
Vue Compatibility: Assumes vue-facing-decorator compatibility (needs validation)
Migration Timing: Assumes migrations run on app startup (automatic execution)
Platform Support: Assumes same behavior across Web (Absurd-SQL) and Mobile (Capacitor SQLite)

Implementation Details

Migration 003: Table Creation

Creates the active_identity table with:

Primary Key: Auto-incrementing ID
Scope Field: For future multi-profile support (currently 'default')
Active DID: Foreign key to accounts.did with CASCADE UPDATE
Timestamps: ISO format timestamps for audit trail
Indexes: Performance optimization for scope and DID lookups

Migration 004: Column Removal

Implements Phase C by:

Table Rebuild: Creates new settings table without activeDid column
Data Preservation: Copies all other data from legacy table
Index Recreation: Rebuilds necessary indexes
Trigger Cleanup: Removes dual-write triggers

Service Façade Implementation

The PlatformServiceMixin extension provides:

Dual-Read Logic: Prefers new table, falls back to legacy during transition
Dual-Write Logic: Updates both tables during Phase A/B
Validation: Ensures DID exists before setting as active
Transaction Safety: Wraps operations in database transactions
Error Handling: Comprehensive logging and error propagation

Feature Flag System

Controls migration phases through:

USE_ACTIVE_IDENTITY_ONLY: Disables legacy fallback reads
DROP_SETTINGS_ACTIVEDID: Enables Phase C column removal
LOG_ACTIVE_ID_FALLBACK: Logs when legacy fallback is used
ENABLE_ACTIVE_IDENTITY_MIGRATION: Master switch for migration system

Security Considerations

Data Validation

DID format validation (basic "did:" prefix check)
Foreign key constraints ensure referential integrity
Transaction wrapping prevents partial updates

Access Control

No row-level security implemented
Scope isolation framework in place for future use
Validation prevents setting non-existent DIDs as active

Audit Trail

Timestamps on all active identity changes
Logging of fallback usage and errors
Migration tracking through built-in system

Performance Characteristics

Read Operations

Primary Path: Single query to active_identity table
Fallback Path: Additional query to settings table (Phase A only)
Indexed Fields: Both scope and active_did are indexed

Write Operations

Dual-Write: Updates both tables during transition (Phase A/B)
Transaction Overhead: All operations wrapped in transactions
Trigger Execution: Additional database operations per update

Migration Impact

Table Creation: Minimal impact (runs once)
Column Removal: Moderate impact (table rebuild required)
Data Seeding: Depends on existing data volume

Testing Strategy

Unit Testing

Service façade method validation
Error handling and edge cases
Transaction rollback scenarios

Integration Testing

Migration execution and rollback
Cross-platform compatibility
Performance under load

End-to-End Testing

Component integration
User workflow validation
Migration scenarios

Deployment Considerations

Rollout Strategy

Phase A: Deploy with dual-write enabled
Phase B: Update components to use new methods
Phase C: Enable column removal (irreversible)

Rollback Plan

Phase A/B: Disable feature flags, revert to legacy methods
Phase C: Requires database restore (no automatic rollback)

Monitoring

Track fallback usage through logging
Monitor migration success rates
Alert on validation failures

Status: Implementation complete, ready for testing and component migration Next Review: After initial testing and component updates Maintainer: Development team

11 KiB Raw Blame History