For large-scale systems handling user online status, UUIDs are often used as unique identifiers. To manage this data efficiently with Redis, consider the following architectural approaches.

1. Using Redis Hashes (Hashes)

Using full UUID strings as direct keys can be inefficient. A better method is to store statuses in a hash, using the UUID as the field and the status as its value.

For example, create a hash named user_activity. A value of 1 indicates online, 0 indicates offline.

# Set a user's status to online
HSET user_activity <USER_UUID> 1

# Retrieve a specific user's status
HGET user_activity <USER_UUID>

# Batch operations for multiple users
HMSET user_activity <UUID_A> 1 <UUID_B> 1
HMGET user_activity <UUID_A> <UUID_B>

2. Using Redis Sets

A alternative is to maintain a set containing the UUIDs of all currently active users. Add UUIDs upon login and remove them upon logout.

# User logs in: add to active set
SADD active_users <USER_UUID>

# User logs out: remove from active set
SREM active_users <USER_UUID>

# Get all currently active user IDs
SMEMBERS active_users

# Get the count of active users
SCARD active_users

# Check if a specific user is active (returns 1 if member, 0 if not)
SISMEMBER active_users <USER_UUID>

3. Data Sharding and Redis Cluster

At a scale of billions, data must be distributed across multiple Redis instances. Implement sharding by using a hash of the UUID to determine the target Redis node.

Redis Cluster automaets this process, handling data partitioning, re-sharding, and failover, simplifying the management of large datasets.

4. Expiration Policies for Memory Management

To prevent memory exhaustion from stale entries, apply an expiration time to each user's status record.

# Set a user's online status with a 30-minute expiration
HSET user_activity <USER_UUID> 1
EXPIRE user_activity <USER_UUID> 1800

# Note: In a hash, you can't expire individual fields. A common pattern is to use separate keys.
# For per-user key expiration:
SET user:status:<USER_UUID> 1 EX 1800

5. Data Persistence and Backup

Configure Redis persistence mechanisms like RDB snapshots or AOF logging to enable data recovery after a restart. Establish regular backup procedures for disaster recovery scenarios.

6. System Monitoring and Alerting

Deploy comprehensive monitoring for key metrics such as memory usage, command latency, and cluster health. Configure alerts to notify operators of performance degradation or failures promptly.