INTRODUCTION REDIS

Maarten Smeets
09-04-2018
 INTRODUCTION REDIS
INTRODUCING REDIS DATATYPES PUBLISH SUBSCRIBE

PERSISTENCE TRANSACTIONS

REJSON AND REDISEARCH CLIENTS

 INTRODUCTION REDIS

1 2 3
What is Redis Why use Redis Who uses Redis
INTRODUCING REDIS

• Redis
First release 2009
An open-source in-memory database project implementing
a distributed, in-memory key-value store with optional
durability.
INTRODUCING REDIS

• Redis is an open source, in-memory data structure store

Can be used as Database, Cache, Message broker

• NoSQL Key/Value store

• Supports multiple data structures

• Features like
• Transactions
• Pub/Sub
• Scalability / availability options
• Time to live for entries Mostly single threaded!
Modules can be multi threaded
TYPICAL USES OF REDIS

• Cache database query results

• https://redislabs.com/ebook/part-1-getting-started/chapter-2-anatomy-of-a-redis-web-application/2-4-
database-row-caching/

• Cache entire webpages (e.g. Wordpress)

• https://wordpress.org/plugins/redis-cache/

• Use for HTTP session store

• https://docs.spring.io/spring-session/docs/current/reference/html5/

• Cache logins / cookies

• https://redislabs.com/ebook/part-1-getting-started/chapter-2-anatomy-of-a-redis-web-application/2-1-
login-and-cookie-caching/
WHY REDIS?

• Very flexible • No schemas, column names

• Very fast • Rich Datatype Support

• Caching & Disk persistence

HOW TO USE REDIS
EXAMPLE: CACHE DB RESULTS
Cache

Clients Back-end / server side

Persistent storage
WHO USES REDIS

Azure Redis Cache

WHO USES REDIS
IN THE NETHERLANDS
 DATATYPES

1 2 3
Strings and expiry Lists, Sets, Hashes Sorted sets
DATATYPES

• Strings • Hashes Datatypes cannot be nested!

E.g. no lists of hashes

You can name your variables like:

• Lists • Bitmaps person:1
person:2

• Sets • Hyperlogs KEYS person:* gives all person variables

The KEYS command is blocking.

Better to keep a set or list of person keys
• Sorted sets • Geospatial
indexes
STRINGS

• Max size 512Mb • Example usage

127.0.0.1:6379> SET greeting

• Byte save. Can store binaries ‘Hello’
OK
127.0.0.1:6379> GET greeting
• Use cases: “Hello”
127.0.0.1:6379> DEL greeting
• Store JPEG’s (integer) 1
• Store serialized objects 127.0.0.1:6379> EXISTS greeting
(integer) 0
EXPIRY

• Expiry can be set for existing variables

127.0.0.1:6379> SET greeting “Hello”

OK
127.0.0.1:6379> EXPIRE greeting 10
(integer) 1
127.0.0.1:6379> TTL greeting
(integer) 8
127.0.0.1:6379> GET greeting
“Hello”
127.0.0.1:6379> GET greeting
(nil)
EXPIRY

• Expiry can be set when a variable is created and expiry can be removed

127.0.0.1:6379> SETEX greeting 10 “Hello”

OK
127.0.0.1:6379> TTL greeting
(integer) 8
127.0.0.1:6379> PERSIST greeting
OK
127.0.0.1:6379> TTL greeting
(integer) -1
LISTS

• A list of Strings 127.0.0.1:6379> LPUSH people “Maarten”

• Max size 4294967295 (integer) 1
127.0.0.1:6379> RPUSH people “John”
(integer) 2
• Can for example be used for 127.0.0.1:6379> LRANGE people 0 -1
• timelines of social networks
1) “Maarten”
2) “John”
• Speed 127.0.0.1:6379> LLEN people
• Actions at the start or end of the list are very fast. (integer) 2
• Actions in the middle are a little less fast 127.0.0.1:6379> LPOP people
“Maarten”
127.0.0.1:6379> RPOP people
“John”
SETS

• Unordered collection of Strings 127.0.0.1:6379> SADD cars “Honda”

(integer) 1
127.0.0.1:6379> SADD cars “Ford”
• Does not allow repeated elements (integer) 1
127.0.0.1:6379> SADD cars “Honda”
(integer) 0
• Useful for tracking unique items 127.0.0.1:6379> SISMEMBER cars “Honda”
(integer) 1
• Allows extracting random members 127.0.0.1:6379> SISMEMBER cars “BMW”
Using SPOP, SRANDMEMBER (integer) 0
127.0.0.1:6379> SMEMBERS cars
1) “Ford”
• Useful for intersects and diffs 2) “Honda”
127.0.0.1:6379> SMOVE cars mycars “Honda”
(integer) 1
127.0.0.1:6379> SDIFF cars mycars
1) “Ford”
HASHES

• Maps between string fields and values 127.0.0.1:6379> HMSET user:1000 username antirez
password P1pp0 age 34
OK
127.0.0.1:6379> HGETALL user:1000
• Ideal for storing objects 1) “username”
2) “antirez”
3) “password”
4) “P1pp0”
5) “age”
6) “34”
127.0.0.1:6379> HSET user:1000 password 12345
(integer) 0
127.0.0.1:6379> HGET user:1000 password
“12345”
127.0.0.1:6379> HKEYS user:1000
1) “username”
2) “password”
3) “age”
SORTED SETS

• Non repeating collections of Strings 127.0.0.1:6379> ZADD myzset 1 "one"

(integer) 1
127.0.0.1:6379> ZADD myzset 1 "one"
• Every member has a score. (integer) 0
127.0.0.1:6379> ZADD myzset 1 "uno"
(integer) 1
• Members are unique. Scores are not 127.0.0.1:6379> ZADD myzset 2 "two" 3 "three"
(integer) 2
127.0.0.1:6379> ZRANGE myzset 0 -1 WITHSCORES
• Ordering is from small to large score 1) "one"
2) "1"
3) "uno"
• Ordering for items with the same score is 4) "1"
alphabetic 5) "two"
6) "2"

• Useful for leader boards or autocomplete 7) "three"

8) "3"
127.0.0.1:6379> ZRANGE myzset 0 0
1) "one"
 PERSISTENCE

1 2
Redis Database File (RDB) Append Only File (OAF)
PERSISTENCE

RDB (Redis Database File) AOF (Append Only File)

Provides point in time snapshots Logs every write

Replays at server startup.

Creates complete snapshot at specified interval
If log gets big, optimization takes place

File is in binary format File is easily readable

On crash minutes of data can be lost Minimal chance of data loss
Small files, fast (mostly) Big files, ‘slow’

https://www.slideshare.net/eugef/redis-persistence-in-practice-1
 PUBLISH SUBSCRIBE

1 2
How to use it Things to mind
PUBLISH SUBSCRIBE
HOW DOES IT WORK Channel can contain
glob patterns like news.*

• SUBSCRIBE [channel]

• UNSUBSCRIBE [channel]

• PUBLISH [channel] [message]

PUBLISH SUBSCRIBE
THINGS TO MIND

• A missed message has been missed permanently.

No retry

• There is no history of messages

Like a JMS topic without durable subscribers
 TRANSACTIONS

1 2
TRANSACTIONS There is no ZPOP but
it can be implemented like below

WATCH zset
• Not like relational database transactions! element = ZRANGE zset 0 0
MULTI
ZREM zset element
• Start a transaction: MULTI EXEC
Commands after MULTI are queued
Redis LUA scripts are also
• Execute the queued commands: EXEC executed in a transaction
All or none of the commands are executed
However, if one fails, the others are still executed

• Make EXEC conditional: WATCH

EXEC will only execute if watched variables are unchanged
 JSON AND SEARCH

1 2
ReJSON RediSearch
REDIS AS A JSON STORE

JSON support can be implemented by adding the ReJSON module from Redis Labs

127.0.0.1:6379> JSON.SET scalar . '"Hello JSON!"'

OK
127.0.0.1:6379> JSON.SET object . '{"foo": "bar", "ans": 42}'
OK
127.0.0.1:6379> JSON.GET object
"{"foo":"bar","ans":42}"
127.0.0.1:6379> JSON.GET object .ans
"42"

https://redislabs.com/blog/redis-as-a-json-store/
 REDISEARCH - REDIS POWERED SEARCH ENGINE

Auto completion
127.0.0.1:6379> FT.SUGADD autocomplete “hello world” 100
OK
127.0.0.1:6379> FT.SUGGET autocomplete “he”
1) "hello world“
Full text search
127.0.0.1:6379> FT.ADD myIdx doc1 1.0 FIELDS title “hello” body “bla” url “http://redis.io”
OK
127.0.0.1:6379> FT.SEARCH myIdx “hello world” LIMIT 0 10
1) (integer) 1
2) “doc1”
3) 1) “title”
2) “hello”
3) “body”
4) “bla”
5) "url”
6) “http://redis.io”
http://redisearch.io/Quick_Start/
docker run -p 6379:6379 redislabs/redisearch:latest
 CLIENTS

1 2
Node.js Spring Boot
CLIENTS
NODE.JS

Create a client and connect

Create an event handler

Subscribe to a channel
CLIENTS
SPRING BOOT

• There is no annotation available to

• Create a container
• Register a listener to the container
• Register a receiver to the listener

• This requires more code than for example listening to a Kafka topic

https://spring.io/guides/gs/messaging-redis/
CLIENTS
SPRING BOOT

Wait for a single message

CLIENTS
SPRING BOOT
SOME THINGS TO MIND IN PRODUCTION

• Redis is single threaded

• Want to use more CPU’s? Use more Redis instances!
• Requests are blocking. Be careful with for example KEYS commands. Mind
the time complexity of operations!

• Mind the number of connections!

• Implement a proxy, for example twemproxy (manages persistent connections)

• Snapshotting is blocking
• Investigate persistence requirements and consider rolling BGSAVE, OAF
instead of snapshotting

http://tech.trivago.com/2017/01/25/learn-redis-the-hard-way-in-production/