How to Elegantly Implement an Online User Count with Redis ZSET

Introduction

Online user counting is a common requirement; the article proposes using Redis sorted set (ZSET) to store active users.

Identifying Online Users

For sites that require login, a valid token indicates that the user is online. For public sites, a unique identifier such as a browser fingerprint (User‑Agent, HTTP headers, screen resolution, timezone, plugins) can be generated using JavaScript libraries like FingerprintJS or ClientJS. The fingerprint is stored in a cookie or header and sent to the backend.

// Install: npm install @fingerprintjs/fingerprintjs
// Usage example:
import FingerprintJS from '@fingerprintjs/fingerprintjs';
FingerprintJS.load().then(fp => {
  fp.get().then(result => {
    const visitorId = result.visitorId;
    console.log(visitorId);
  });
});

Implementation Steps

1. Add online user with ZADD

The zadd command takes three arguments: key (the ZSET name), score (a numeric value, here the expiration timestamp), and member (the user identifier).

Example of adding a simple member: ZADD myzset 1 "one" In practice the code adds a user token with an expiration time as the score, ensuring each user appears only once with the latest login state:

LocalDateTime expireTime = LocalDateTime.now().plusSeconds(expireTimeout);
String expireTimeStr = DateUtil.formatFullTime(expireTime);
redisService.zadd("user.active", Double.parseDouble(expireTimeStr), userToken);

2. Query online count with ZRANGEBYSCORE

zrangeByScore

returns members whose scores fall within a given range. To obtain all currently active users, query from the current timestamp to +inf:

String now = DateUtil.formatFullTime(LocalDateTime.now());
Set<String> userOnlineStringSet = redisService.zrangeByScore("user.active", now, "+inf");
int onlineCount = userOnlineStringSet.size();

3. Remove expired entries with ZREMRANGEBYSCORE

A scheduled task deletes members whose scores are older than the current time, effectively clearing offline users:

String now = DateUtil.formatFullTime(LocalDateTime.now());
redisService.zremrangeByScore("user.active", "-inf", now);

4. Remove a user on logout with ZREM

When a user explicitly logs out, remove the corresponding member from the ZSET:

redisService.zrem("user.active", "xxx");

Conclusion

The core logic is to maintain a Redis ZSET where the member is the user identifier and the score is the expiration timestamp. Adding, querying, and cleaning the set with zadd, zrangeByScore, zremrangeByScore and zrem provides a simple, efficient way to count online users.