Configure Database Replication

Time: 10-15 minutes
Prerequisites:

Redis Enterprise cluster with multiple nodes
redisctl configured with Enterprise credentials
Database already created

What is Replication?

Replication provides:

High availability - automatic failover if master fails
Read scalability - distribute reads across replicas
Data durability - multiple copies of data

Quick Setup

# Enable replication on existing database
redisctl enterprise database update \
  --database-id 1 \
  --data '{
    "replication": true,
    "shards_count": 2
  }' \
  --wait

Step-by-Step Setup

1. Create Database with Replication

redisctl enterprise database create \
  --data '{
    "name": "replicated-db",
    "memory_size": 1073741824,
    "type": "redis",
    "port": 12000,
    "replication": true,
    "shards_count": 1,
    "sharding": false
  }' \
  --wait

2. Verify Replication Status

redisctl enterprise database get \
  --database-id 1 \
  -o json \
  -q '{
    name: name,
    replication: replication,
    shards_count: shards_count,
    endpoints: endpoints
  }'

3. Check Shard Distribution

redisctl enterprise shard list-by-database \
  --bdb-uid 1 \
  -o json \
  -q '[].{
    uid: uid,
    role: role,
    node: node_uid,
    status: status
  }'

Expected: One master and one replica shard on different nodes.

Replication Topology

Single Master with Replica

# Default configuration
{
  "replication": true,
  "shards_count": 1,
  "sharding": false
}
# Result: 1 master + 1 replica = 2 total shards

Sharded with Replication

# Clustered database with replication
{
  "replication": true,
  "shards_count": 3,
  "sharding": true
}
# Result: 3 master + 3 replica = 6 total shards

Advanced Configuration

Set Replica Count

# Multiple replicas per master
redisctl enterprise database update \
  --database-id 1 \
  --data '{
    "replication": true,
    "replica_sources": [
      {"replica_source_name": "replica1", "replica_source_type": "replica"},
      {"replica_source_name": "replica2", "replica_source_type": "replica"}
    ]
  }' \
  --wait

Rack Awareness

Ensure master and replicas are on different racks/zones:

redisctl enterprise database update \
  --database-id 1 \
  --data '{
    "rack_aware": true
  }' \
  --wait

Monitoring Replication

Check Replication Lag

# Get replication lag for database
redis-cli -h localhost -p 12000 INFO replication

# Or via REST API
redisctl enterprise database get \
  --database-id 1 \
  -q 'replica_sync[].{
    replica: replica_uid,
    lag: lag,
    status: status
  }'

Monitor Sync Status

# Check if replicas are in sync
redisctl enterprise shard list-by-database \
  --bdb-uid 1 \
  -o json \
  -q '[?role==`replica`].{
    shard: uid,
    status: status,
    sync_status: sync_status
  }'

Failover Operations

Manual Failover

# Failover specific shard
redisctl enterprise shard failover \
  --uid 1:1 \
  --force

# Verify new master
redisctl enterprise shard get --uid 1:1 -q 'role'

Automatic Failover

Enabled by default with replication:

# Check failover settings
redisctl enterprise database get \
  --database-id 1 \
  -q '{
    replication: replication,
    watchdog_profile: watchdog_profile
  }'

Replica Configuration

Read-Only Replicas

# Configure replica as read-only (default)
redisctl enterprise database update \
  --database-id 1 \
  --data '{
    "replica_of": {
      "endpoints": ["master-db:12000"],
      "readonly": true
    }
  }' \
  --wait

External Replication Source

Replicate from external Redis:

redisctl enterprise database create \
  --data '{
    "name": "replica-db",
    "memory_size": 1073741824,
    "type": "redis",
    "port": 12001,
    "replica_of": {
      "endpoints": ["external-redis.example.com:6379"],
      "authentication_redis_pass": "source-password"
    }
  }' \
  --wait

Replication Performance

Optimize Replication Speed

# Increase replication buffer
redisctl enterprise database update \
  --database-id 1 \
  --data '{
    "repl_backlog_size": 104857600  # 100MB
  }' \
  --wait

Monitor Replication Traffic

redisctl enterprise database get \
  --database-id 1 \
  -o json \
  -q '{
    replication_traffic: repl_traffic,
    backlog_size: repl_backlog_size
  }'

Common Patterns

High Availability Setup

# Production-ready HA configuration
redisctl enterprise database create \
  --data '{
    "name": "ha-database",
    "memory_size": 10737418240,
    "type": "redis",
    "port": 12000,
    "replication": true,
    "shards_count": 3,
    "sharding": true,
    "rack_aware": true,
    "data_persistence": "aof",
    "aof_policy": "appendfsync-every-sec"
  }' \
  --wait

Read Scaling with Replicas

# Application pattern: writes to master, reads from replicas
from redis import Redis, Sentinel

# Connect to master for writes
master = Redis(host='master-endpoint', port=12000)
master.set('key', 'value')

# Connect to replica for reads
replica = Redis(host='replica-endpoint', port=12001)
value = replica.get('key')

Disaster Recovery

Backup Replication Status

# Save replication configuration
redisctl enterprise database get \
  --database-id 1 \
  -o json > db-replication-config.json

Restore After Failure

# Recreate database with same configuration
redisctl enterprise database create \
  --data @db-replication-config.json \
  --wait

Common Issues

Replication Lag Increasing

# Check network between nodes
redisctl enterprise node list -o table

# Check shard placement
redisctl enterprise shard list-by-database --bdb-uid 1 -o table

# Consider adding more replicas or increasing bandwidth

Replica Out of Sync

# Force resync
redisctl enterprise shard failover --uid 1:2 --force

# Check sync status
redisctl enterprise shard get --uid 1:2 -q 'sync_status'

Split Brain Scenario

Prevention:

Always use odd number of cluster nodes
Enable rack awareness
Monitor node connectivity

Recovery:

# Identify correct master
redisctl enterprise shard list-by-database --bdb-uid 1 \
  -q '[?role==`master`]'

# Force failover if needed
redisctl enterprise database update \
  --database-id 1 \
  --data '{"action": "recover"}' \
  --wait

Best Practices

Always Enable for Production - Replication is critical for HA
Use Rack Awareness - Distribute across failure domains
Monitor Replication Lag - Alert on high lag
Test Failover - Regularly test automatic failover
Plan Capacity - Replicas consume same resources as master
Persist Configuration - Backup replication settings

Next Steps

Cluster Health Check - Monitor replication health
Node Management - Manage replica placement
Generate Support Package - Troubleshooting tools
Create Database - Database configuration basics

redisctl Documentation