# MCP Gateway v0.9.0 - Edge Cases and Error Conditions
# Edge case testing and error handling validation
# Focus: Boundary conditions, error scenarios, and recovery
worksheet_name: "Edge Cases"
description: "Edge case testing including error conditions, boundary values, and recovery scenarios"
priority: "MEDIUM"
estimated_time: "60-90 minutes"
headers:
- "Test ID"
- "Edge Case Category"
- "Scenario"
- "Test Steps"
- "Expected Behavior"
- "Actual Behavior"
- "Recovery Method"
- "Status"
- "Tester"
- "Date"
- "Severity"
- "Comments"
tests:
- test_id: "EDGE-001"
category: "Empty Database"
scenario: "Fresh installation on empty database"
steps: |
1. Delete existing database file
2. Run migration: python3 -m mcpgateway.bootstrap_db
3. Check system initialization
4. Verify admin user and team creation
expected: "System initializes correctly from completely empty state"
recovery: "Bootstrap migration process"
severity: "Low"
- test_id: "EDGE-002"
category: "Network Interruption"
scenario: "Network connection lost during operation"
steps: |
1. Start long-running operation (large export)
2. Disconnect network interface
3. Wait 30 seconds
4. Reconnect network
5. Check operation recovery
expected: "Graceful error handling, operation retry or proper failure"
recovery: "Retry mechanism or user notification"
severity: "Medium"
- test_id: "EDGE-003"
category: "Orphaned Resources"
scenario: "Resources without team assignments"
steps: |
1. Manually set team_id to NULL: UPDATE tools SET team_id = NULL WHERE id = 'test-id';
2. Navigate to admin UI tools section
3. Check tool visibility
4. Run fix script: python3 scripts/fix_multitenancy_0_7_0_resources.py
5. Verify resource assignment
expected: "Fix script successfully assigns orphaned resources to admin team"
recovery: "Fix script execution"
severity: "High"
- test_id: "EDGE-004"
category: "Large Payloads"
scenario: "Oversized request payloads"
steps: |
1. Create very large JSON payload (>10MB)
2. Send to tool creation endpoint
3. Check request size limits enforced
4. Verify proper error handling
expected: "Request size limits enforced gracefully"
recovery: "Error message with size limit info"
severity: "Medium"
- test_id: "EDGE-005"
category: "Malformed Data"
scenario: "Invalid JSON and parameter formats"
steps: |
1. Send malformed JSON to API endpoints
2. Send invalid parameter types
3. Test with missing required fields
4. Check validation error responses
expected: "Input validation rejects malformed data with helpful errors"
recovery: "Validation error messages"
severity: "Medium"
- test_id: "EDGE-006"
category: "Resource Conflicts"
scenario: "Name conflicts and duplicate identifiers"
steps: |
1. Create team with existing name
2. Try creating tool with existing name in same team
3. Test unique constraint enforcement
4. Check conflict resolution
expected: "Unique constraints enforced, conflicts handled gracefully"
recovery: "Conflict error messages"
severity: "Medium"
- test_id: "EDGE-007"
category: "Session Management"
scenario: "Session expiry and timeout handling"
steps: |
1. Login and get JWT token
2. Wait for token expiry (or manually expire)
3. Try using expired token
4. Test session refresh workflow
expected: "Expired tokens rejected, refresh workflow available"
recovery: "Token refresh or re-authentication"
severity: "Medium"
- test_id: "EDGE-008"
category: "Database Connection"
scenario: "Database becomes unavailable"
steps: |
1. Start gateway normally
2. Stop database service
3. Try API operations
4. Restart database
5. Check connection recovery
expected: "Graceful error handling, automatic reconnection"
recovery: "Connection pool recovery"
severity: "High"
- test_id: "EDGE-009"
category: "Disk Space"
scenario: "Insufficient disk space"
steps: |
1. Fill disk space (test environment only)
2. Try creating resources
3. Try database operations
4. Check error handling
expected: "Disk space errors handled gracefully"
recovery: "Clear error messages"
severity: "Medium"
- test_id: "EDGE-010"
category: "Unicode and Special Characters"
scenario: "International characters and special symbols"
steps: |
1. Create team with Unicode name: 测试团队
2. Create tool with emoji: 🔧 Test Tool
3. Test special characters in descriptions
4. Verify proper encoding/decoding
expected: "Unicode and special characters handled correctly"
recovery: "UTF-8 encoding support"
severity: "Low"
- test_id: "EDGE-011"
category: "Rapid State Changes"
scenario: "Quick successive operations on same resource"
steps: |
1. Create tool
2. Rapidly update tool multiple times
3. Delete and recreate quickly
4. Check state consistency
expected: "State consistency maintained, no race conditions"
recovery: "Locking mechanisms"
severity: "Medium"
- test_id: "EDGE-012"
category: "Migration Interruption"
scenario: "Migration process fails or is interrupted"
steps: |
1. Start migration
2. Interrupt process (Ctrl+C)
3. Check database state
4. Try re-running migration
5. Verify recovery
expected: "Migration can be safely restarted or rolled back"
recovery: "Migration rollback or resume"
severity: "Critical"
- test_id: "EDGE-013"
category: "Team Limits"
scenario: "Exceeding team member or resource limits"
steps: |
1. Set low team limits in configuration
2. Try exceeding member limits
3. Try exceeding resource limits
4. Check limit enforcement
expected: "Limits enforced with clear error messages"
recovery: "Quota management interface"
severity: "Medium"
- test_id: "EDGE-014"
category: "Cross-Database Migration"
scenario: "Migrating data between SQLite and PostgreSQL"
steps: |
1. Setup data in SQLite
2. Export configuration
3. Switch to PostgreSQL
4. Run migration
5. Import configuration
6. Verify data integrity
expected: "Data migrates correctly between database types"
recovery: "Export/import workflow"
severity: "High"
- test_id: "EDGE-015"
category: "Clock Skew"
scenario: "Time synchronization issues"
steps: |
1. Change system clock
2. Test token expiration
3. Test audit logging timestamps
4. Check time-based operations
expected: "Time-based operations handle clock differences gracefully"
recovery: "UTC normalization"
severity: "Low"
