02 Redis

Redis is an advanced key-value store, where keys can contain data structures such as strings, hashes, lists, sets, and sorted sets. Supporting a set of atomic operations on these data types.

Redis is not a replacement for Relational Databases nor Document Stores. It might be used complementary to a SQL relational store, and/or NoSQL document store.

It’s performance and ease of use were made possible by simplifying its architecture as much as possible:

• Written in C
• Single Threaded
• Atomic operations only
• No official support for Windows
• Executes most commands with costant O(1) complexity
• Based on 2 fundamentals commands: SET(k,v) ,which sets the value v to the key, v and GET(k), which returns the value associated to the key k
• In Memory in DB

These characteristics make Redis very fast on reads, but not very efficient for general purpose long term storage system (since being in-memory means that a single shutdown will cost you all your data).

Redis is the best technology for the use cases where:

• Speed is critical
• You need fast access to complex data structures
• Dataset can fit in memory
• Data is not critical

Data types

Redis topologies

Standalone

• The master data is optionally replicated to slaves.
• The slaves provides data redundancy, reads offloading and save-to-disk offloading.
• Clients can connect to the Master for read/write operations or to the Slaves for read operations.
• Slaves can also replicate to its own slaves.
• There is no automatic failover.

Sentinel

Redis Sentinel provides a reliable automatic failover in a master/slave topology, automatically promoting a slave to master if the existing master fails. Sentinel does not distribute data across nodes.

Twemproxy

Twemproxy (from Twitter, not part of Redis) works as a proxy between the clients and many Redis instances. It is able to automatically distribute data among different standalone Redis instances and supports consistent hashing with different strategies and hashing functions. However Multi-key commands and transactions are not supported.

Cluster

Redis Cluster distributes data across different Redis instances and perform automatic failover if any problem happens to any master instance.

• All nodes are directly connected with a service channel.
• The keyspace is divided into hash slots. Different nodes will hold a subset of hash slots.
• Multi-key commands are only allowed for keys in the same hash slot.

Scaling Redis

• Persistence: while all main operations are done in main memory, Redis provides a couple of features to implement disk persistence. The first one allows you to take a snapshot of the db and save it to a file (RDB), while the second one works by keeping track of changes in append only log file (AOF)
• A Redis instance known as the master, ensures that one or more instances kwown as the slaves, become exact copies of the master. Clients can connect to the master or to the slaves. Slaves are read only by default.
• Partitioning: Breaking up data and distributing it across different hosts in a cluster ;can be implemented in different layers:
  • Client: Partitioning on client-side code.
  • Proxy: An extra layer that proxies all redis queries and performs parti- tioning
  • Query Router: instances will make sure to forward the query to the right node. (i.e Redis Cluster).
• Failover: it can be either Manual, Automatic with Redis Sentinel (for master-slave topology) or Automatic with Redis Cluster (for cluster topol- ogy)

Commands

Other than the basic GET and SET commands there are a few other utility functions, but they only build upon the 2 original ones. There are commands to delete and check existence of a key and get the type of a value, for example.

Another important feature is the ability to set an expiration time to a key ,after which the key it’s deleted from the db.