Video a4yX7RUgTxI

[00:00] In this video, we discover the versatility of Redis.

[00:12] We'll talk about the top use cases that have been better tested in production at various companies and at various scales. Get insights into how Redis solves interesting scalability challenges for us

[00:24] and learn why it is a great tool to know well in our system design toolset. Let's dive right into it. First of all, what is Redis and why do people use it?

[00:36] Redis is an in-memory data structure store. It is most commonly used as a cache. It supports many data structures such as strings, hashes, lists, sets, and sort of sets.

[00:51] Redis is known for its speed. We made a video to explain why Redis is so fast. Look for the link in the description if you would like to watch that video. Now let's dive into the top use cases of Redis.

[01:04] The number one use case of Redis is caching objects to speed up web applications. In this use case, Redis stores frequently requested data in memory. It allows the web servers to return frequently accessed data quickly.

[01:19] This reduces the load on the database and improves the response time for the application. At scale, the cache is distributed among a cluster of Redis servers.

[01:31] Charting is a common technique to distribute and caching load evenly across the cluster Other topics to consider when deploying Redis as a distributed cache include setting a correct TTL and handling thundering hurt on cold start.

[01:50] Another common use case is to use Redis as a session store to share session data among stateless servers. When a user logs into a web application, The session data is stored in Redis,

[02:03] along with a unique session ID that is returned to the client as a cookie. When the user makes a request to the application, the session ID is included in the request, and the Stalus web server retrieves the session data from Redis using the ID.

[02:19] It is important to note that Redis is an in-memory database. The session data stored in Redis will be lost if the Redis server restarts. Even though Redis provides persistent options like a snapshot and AOF or append-only file,

[02:36] they allow session data to be saved to disk and reloaded into memory in the event of a restart. These options often take too long to load to be practical. In production, replication is usually used instead.

[02:50] In this case, data is replicated to a backup instance. In the event of a crash of the main instance, the backup instance is quickly promoted to take over the traffic.

[03:02] The next use case is DistributorLock. DistributorLocks are used when multiple nodes in an application need to coordinate access to some shared resource Redis is used as a distributor log with its atomic commands like setNX or setIfNotExists

[03:21] It allows a caller to set a key only if it does not already exist. Here's how it works at a high level. Client1 tries to acquire the log by setting a key with a unique value

[03:34] and a timeout using the setNX command. If the key was not already set, the setNX command returns 1, indicating that the log has been acquired by client1.

[03:46] Client1 finishes its work and releases the log by deleting the key. Now if the key was already set, the setNX command returns 0, indicating that the log

[03:58] has already been held by another client. In this case, client1 waits and retries the set-annex operation until the log is released by the other client. Note that this simple implementation might be good enough for many use cases, but it is not completely fault-tolerant.

[04:16] For production use, there are many Redis client libraries that provide high-quality distributed log implementation built right off the box. Next up is RayLimiter.

[04:28] Redis can be used as a rate limiter by using its increment command on some counters and setting expiration times on those counters. A very basic rate limiting algorithm works this way.

[04:42] For each incoming request the request IP or user ID is used as a key The number of requests for the key is incremented using the increment command in Redis The current count is compared to the allowed rate limit If the count is

[04:57] within the rate limit, the request is processed. If the count is over the limit, the request is rejected. The keys are set to expire after a specific time window, for example a minute,

[05:09] to reset the cans for the next time window. More sophisticated rate limiters, like the leaky bucket algorithm, can also be implemented using Redis.

[05:22] The last use case we would like to talk about is gaming leaderboard. For most games that are not super high scale, Redis is a delightful way to implement various types of gaming leaderboards.

[05:34] Sorta sets are the fundamental data structure that enables this. Assorted set is a collection of unique elements, each with a score associated with it. The elements are sorted by score.

[05:47] This allows for quick retrieval of the elements by score in logarithmic time. That's it for our top Redis use cases. Redis is very versatile. There are many different ways to use it.

[06:00] Some features are more better tested than others. Our creativity is the limit. Use it with care and have fun with it. You would like our videos. You may like our weekly system design newsletter as well.

[06:14] It covers topics and trends in large-scale system design, trusted by 200,000 readers. Subscribe at blog.buybygo.com.