jknapp 7869b81f27 CRITICAL FIX: Migrate HAProxy IP blocking from ACL to map files ...

**Problem Solved:**
- HAProxy ACL 64-word limit caused config parsing failures
- "too many words, truncating after word 64" error
- Complete service outage when >64 IPs were blocked
- Error: "no such ACL : 'is_blocked'" broke all traffic routing

**Solution: HAProxy Map Files (v1.6+)**
- ✅ Unlimited IP addresses (no word limits)
- ✅ Runtime updates without config reloads
- ✅ Better performance (hash table vs linear search)
- ✅ Safer config management with validation & rollback

**Technical Implementation:**

**Map File Integration:**
- `/etc/haproxy/blocked_ips.map` stores all blocked IPs
- `http-request deny status 403 if { src -f /etc/haproxy/blocked_ips.map }`
- Runtime updates: `echo "add map #0 IP" | socat stdio /var/run/haproxy.sock`

**Safety Features Added:**
- `create_backup()` - Automatic config/map backups before changes
- `validate_haproxy_config()` - Config validation before applying
- `restore_backup()` - Automatic rollback on failures
- `reload_haproxy_safely()` - Safe reload with validation pipeline

**Runtime Management:**
- `update_blocked_ips_map()` - Sync database to map file
- `add_ip_to_runtime_map()` - Immediate IP blocking without reload
- `remove_ip_from_runtime_map()` - Immediate IP unblocking

**New API Endpoints:**
- `POST /api/config/reload` - Safe config reload with rollback
- `POST /api/blocked-ips/sync` - Sync database to runtime map

**Template Changes:**
- Replaced ACL method: `acl is_blocked src IP1 IP2...` (64 limit)
- With map method: `http-request deny if { src -f blocked_ips.map }` (unlimited)

**Backwards Compatibility:**
- Existing API endpoints unchanged (GET/POST/DELETE /api/blocked-ips)
- Database schema unchanged
- Automatic migration on first config generation

**Performance Improvements:**
- O(1) hash table lookups vs O(n) linear ACL search
- No config reloads needed for IP changes
- Supports millions of IPs if needed
- Memory efficient external file storage

**Documentation:**
- Complete migration guide in MIGRATION_GUIDE.md
- Updated API documentation with new endpoints
- Runtime management examples
- Troubleshooting guide

**Production Safety:**
- All changes include automatic backup/restore
- Config validation prevents bad deployments
- Runtime updates avoid service interruption
- Comprehensive error logging and monitoring

This fixes the critical production outage caused by ACL word limits
while providing a more scalable and performant IP blocking solution.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

2025-08-22 08:31:17 -07:00

33 KiB

Raw History

View Raw