Redis for DBAs: Quick Installation, Core Commands, Caching Design Patterns, and High‑Concurrency Practices

Redis is gaining popularity among DBAs because its in‑memory read/write speed can handle high‑concurrency scenarios that traditional relational databases struggle with.

Installation (Aimalinux version)

# 下载+编译安装
wget http://download.redis.io/releases/redis-7.0.0.tar.gz

tar xzf redis-7.0.0.tar.gz

cd redis-7.0.0

make && make install

# 启动Redis服务（后台运行）
nohup src/redis-server &

Default port is 6379; remember to open the firewall.

Essential Redis commands

# 存数据（String类型）
SET user:1001 "张三"

# 取数据
GET user:1001

# 存哈希表（适合存对象）
HSET product:8888 name "手机" price 3999

# 设置过期时间（30秒后自动删除）
EXPIRE product:8888 30

# 查看内存信息
INFO memory

Cache design pitfalls and solutions

Cache penetration : use empty‑value caching or Bloom filters.

Cache avalanche : assign random expiration times, e.g., 300 ± 60 seconds .

Hot key overload : combine local cache (Guava) with Redis replicas.

// Java example: cache empty value to prevent penetration
public String getProduct(String id) {
    String key = "product:" + id;
    String value = redis.get(key);
    if (value == null) {
        value = db.query("SELECT * FROM product WHERE id=?", id);
        if (value == null) {
            redis.setex(key, 300, ""); // cache empty for 5 minutes
        } else {
            redis.setex(key, 3600, value); // cache normal data for 1 hour
        }
    }
    return value;
}

Redis data structures (Java examples)

String – store numbers, JSON, or binary data. Jedis jedis = new Jedis("localhost"); jedis.set("user:1001", "{'name':'老王','age':28}"); String user = jedis.get("user:1001");

Hash – store object attributes. Map product = new HashMap<>(); product.put("price", "99.99"); product.put("stock", "1000"); jedis.hset("product:888", product);

List – implement queues such as order processing. jedis.lpush("order:queue", "order1001", "order1002"); String latestOrder = jedis.rpop("order:queue");

Set – deduplication and UV statistics. redis-cli SADD active_users:20240507 "user1001" "user1002" redis-cli SCARD active_users:20240507

ZSet – ranking, e.g., game leaderboards. jedis.zadd("game_rank", 5000, "player1"); jedis.zadd("game_rank", 8000, "player2"); Set top3 = jedis.zrevrange("game_rank", 0, 2);

High‑concurrency patterns (Java + Redis)

1. Distributed lock using SETNX (or SET … NX EX ) to prevent overselling.

public boolean lock(String key, String value, int expireTime) {
    String result = jedis.set(key, value, "NX", "EX", expireTime);
    return "OK".equals(result);
}

// Unlock with Lua script for atomicity
String script = "if redis.call('get', KEYS[1]) == ARGV[1] then return redis.call('del', KEYS[1]) else return 0 end";
jedis.eval(script, Collections.singletonList(key), Collections.singletonList(value));

2. Flash‑sale (秒杀) using a Lua script to atomically decrement stock.

-- Lua script (atomic stock decrement)
local stock = tonumber(redis.call('GET', KEYS[1]))
if stock > 0 then
    redis.call('DECR', KEYS[1])
    return 1   -- success
else
    return 0   -- failure
end

Pre‑warm stock: SET stock:1001 1000 .

Redis cluster deployment

1. Master‑slave replication (one master, two slaves) – configure redis.conf with bind 0.0.0.0 , port 6379 , and replicaof for slaves.

2. Sentinel for automatic failover – sentinel.conf example:

sentinel monitor mymaster
6379 2
sentinel down-after-milliseconds mymaster 5000
sentinel failover-timeout mymaster 60000

3. Cluster mode for sharding – create a 6‑node (3 master, 3 replica) cluster:

redis-cli --cluster create node1:6379 node2:6379 node3:6379 node4:6379 node5:6379 node6:6379 --cluster-replicas 1

Operations essentials

Persistence: RDB snapshots and AOF logs (enable with appendonly yes ).

Monitoring: redis-cli info stats for ops/sec, redis-cli info memory for memory usage.

Connection pool tuning (Java):

JedisPoolConfig config = new JedisPoolConfig();
config.setMaxTotal(100); // max connections
config.setMaxIdle(20);   // max idle
JedisPool pool = new JedisPool(config, "redis-host", 6379);

Mastering Redis enables DBAs and developers to dramatically improve database performance and handle high‑traffic workloads.