Skip to content
advanced Phase 81 · HA Fundamentals

Redundancy

Redundancy patterns including active-active, active-passive, database redundancy, and storage redundancy

45m
0 problems
Topic Progress 0%

Active-Active vs Active-Passive

Architecture Comparison

Active-Active:                    Active-Passive:
┌─────────┐ ┌─────────┐        ┌─────────┐ ┌─────────┐
│ Node 1  │ │ Node 2  │        │ Active  │ │Standby  │
│(Active) │ │(Active) │        │  Node   │ │  Node   │
└────┬────┘ └────┬────┘        └────┬────┘ └────┬────┘
     │           │                   │           │
     └─────┬─────┘                   └─────┬─────┘
           â–¼                               â–¼
      Both serving               Only active serving
      Traffic                    (failover on failure)

Comparison

Aspect Active-Active Active-Passive
Resource usage 100% both 50% standby
Failover Instant Minutes
Complexity Higher Lower
Cost 2x 1.5x
Load sharing Yes No

Magento Configuration

// Active-Active: Load balancer distributes to both
// app/etc/env.php
'cache' => [
    'frontend' => [
        'default' => [
            'backend' => 'Magento\Framework\Cache\Backend\Redis',
            'backend_options' => [
                'server' => 'redis-cluster.example.com', // Shared Redis
                'port' => '6379'
            ]
        ]
    ]
],
// Both web nodes use same shared Redis/DB

Database Redundancy

MySQL Replication Setup

Primary (Write) → Replica 1 (Read)
                → Replica 2 (Read)
                → Replica 3 (Read)

All replicas sync from primary
Read queries distributed across replicas

Galera Cluster (Multi-Primary)

┌─────────┐     ┌─────────┐     ┌─────────┐
│ Node 1  │◀───▶│ Node 2  │◀───▶│ Node 3  │
│(Primary)│     │(Primary)│     │(Primary)│
└─────────┘     └─────────┘     └─────────┘

Synchronous replication
Any node can handle writes
Automatic conflict resolution

Database Redundancy Config

// app/etc/env.php
'db' => [
    'connection' => [
        'default' => [
            'host' => 'primary-db.example.com',
            'dbname' => 'magento',
            'username' => 'magento_rw',
            'password' => '***',
            'active' => '1'
        ],
        'read' => [
            ['host' => 'replica1-db.example.com'],
            ['host' => 'replica2-db.example.com'],
            ['host' => 'replica3-db.example.com']
        ]
    ]
],

Storage Redundancy

Storage Options

Option          | Redundancy     | Performance | Cost
────────────────|────────────────|────────────|──────
Local RAID      | Disk-level     | High        | Low
EBS volumes     | AZ-level       | High        | Medium
EFS/NFS         | Regional       | Medium      | Medium
S3              | Regional       | Medium      | Low
Replicated SAN  | Site-level     | High        | High

File System Redundancy

# RAID 10 for Magento filesystem
mdadm --create /dev/md0 --level=10 --raid-devices=4 \
    /dev/sdb1 /dev/sdc1 /dev/sdd1 /dev/sde1

# Mount for Magento
mount /dev/md0 /var/www/magento

Media Storage Redundancy

// Use S3-compatible storage for media
// app/etc/env.php
'filesystem' => [
    'drivers' => [
        'aws' => [
            'config' => [
                'bucket' => 'magento-media-prod',
                'region' => 'us-east-1',
                'use_path_style_endpoint' => false
            ]
        ]
    ]
],
// S3 provides 99.999999999% durability

Redundancy Architecture

Full Redundancy Design

┌─────────────────────────────────────────┐
│           Load Balancer (2 nodes)        │
└────────────────┬────────────────────────┘
                 │
┌────────────────┴────────────────────────┐
│     Web Tier (Active-Active, 3 nodes)    │
└────────────────┬────────────────────────┘
                 │
┌────────────────┴────────────────────────┐
│    Application Tier (Shared Redis)       │
└────────────────┬────────────────────────┘
                 │
┌────────────────┴────────────────────────┐
│     Database Tier (Primary + 2 Replicas) │
└────────────────┬────────────────────────┘
                 │
┌────────────────┴────────────────────────┐
│     Storage Tier (RAID + S3 Backup)      │
└─────────────────────────────────────────┘

Redundancy Checklist

1. Load balancer: 2+ nodes with VIP
2. Web nodes: 3+ with shared sessions
3. Database: Primary + 2 replicas minimum
4. Cache: Redis cluster with 3+ nodes
5. Search: 3+ OpenSearch nodes
6. Storage: RAID + offsite backup
7. Network: Multiple network paths

Quiz

1. What is the main advantage of active-active over active-passive?

Question 1 options

2. How many database replicas are recommended for redundancy?

Question 2 options

3. What does S3 provide for storage redundancy?

Question 3 options

Flashcards

Question

Active-active vs active-passive?

Click to reveal answer

Answer

Active-active: both nodes serve traffic, instant failover; Active-passive: standby waits for failover

Question

Database redundancy minimum?

Click to reveal answer

Answer

Primary + 2 replicas for HA and read scaling

Question

S3 durability?

Click to reveal answer

Answer

99.999999999% (11 nines)

Question

Full redundancy checklist?

Click to reveal answer

Answer

LB, web nodes, DB, cache, search, storage, network

Revision Notes

Key Takeaways

  • 1. Active-active provides instant failover and load sharing
  • 2. Active-passive is simpler but has failover delay
  • 3. Database needs primary + 2 replicas minimum for HA
  • 4. S3 provides 11 nines durability for storage
  • 5. Every layer needs redundancy to eliminate SPoFs

Interview Tips

  • • Compare active-active vs active-passive trade-offs
  • • Explain database replication and redundancy strategies
  • • Design redundant architecture for all components

Cheat Sheet

Redundancy:
  Active-Active: Both serve, instant failover
  Active-Passive: Standby waits, delayed failover

Database:
  Primary + 2+ replicas
  Galera for multi-primary

Storage:
  RAID for local redundancy
  S3 for offsite (11 nines durability)

Architecture:
  LB (2+), Web (3+), DB (1+2), Redis (3+)
  Every layer redundant