Radish set against hash

Question

In many Redis textbooks (such as this one ), data is stored in a set, but with several values ​​combined into a string (that is, a user account can be stored in a set as two entries: "user: 1000: username" and "user: 1000: password ").

However, Redis also has hashes. It seems like it makes sense to have a hash of "user: 1000" that contains the entry "username" and the entry "password". Instead of concatenating strings to access a specific value, you simply access them directly in the hash.

So why is it not used so much? Is this just old textbooks? Or do Redis hashes have performance issues?

Answer 1

Redis hashes are good for storing more complex data, as you suggest in your question. I use them just for this - to store objects with several attributes that need to be cached (in particular, inventory data for a specific product on an e-commerce site). Of course, I could use a concatenated string, but this adds unnecessary complexity to my client code, and updating a single field is not possible.

Perhaps you are right - study guides may simply be due to the fact that hashes were introduced. They were clearly designed to store object representations: http://oldblog.antirez.com/post/redis-weekly-update-1.html

I believe that one of the problems would be the number of commands that Redis should serve when a new element is inserted (n is the number of commands, where n is the number of fields in the Hash) compared to the simple String SET command. I did not find that this is a problem yet on a service that hits Redis about 1 million times a day. Using the right data structure is more important to me than a slight performance impact.

(Also see my comment regarding Redis Sets vs. Redis Strings - I think your question relates to strings, but correct me if I am wrong!)

Answer 2

Hashes are one of the most efficient storage methods in Redis, even if they allow you to recommend them for use whenever possible.

http://redis.io/topics/memory-optimization

Use hashes if possible

Small hashes are encoded in a very small space, so you should try to present your data with hashes whenever possible. For example, if you have objects representing users in a web application, instead of using different keys for your first name, last name, email, password, use one hash with all the necessary fields.