Redis is a key-value store that can be used as a database, cache, and message broker. It supports basic data structures like strings, hashes, lists, sets, sorted sets with operations that are fast thanks to storing the entire dataset in memory. Redis also provides features like replication, transactions, pub/sub messaging and can be used for caching, queueing, statistics and inter-process communication.

1. More than just a key value datastore
Christopher Spring

2. What is redis?

3. remote dictionary server

4. remote dictionary server
'Key' => String
'Key' => Set
'Key' => Hash
'Key' => List
'Key' => Sorted Set

5. Key features

6. Key features

7. Key features
Cheap API for lists, sets...

8. Key features
Cheap API for lists, sets...
Atomic operations

9. Key features
Cheap API for lists, sets...
Atomic operations
Transactions

10. Key features
Cheap API for lists, sets...
Atomic operations
Transactions
Key expiry

11. In memory storage

12. In memory storage
Single process event driven design

13. In memory storage
Single process event driven design
Non blocking

14. In memory storage
Single process event driven design
Non blocking
100k+ operations per second

15. Asynchronous persistence

16. Asynchronous persistence
Snapshotting
save 60 1000

17. Asynchronous persistence
Snapshotting
save 60 1000
Append only file
Log rewriting
fsync

18. Master slave replication
slaveof '192.168.1.1' 6379

19. Master slave replication
slaveof '192.168.1.1' 6379
Multiple slaves

20. Master slave replication
slaveof '192.168.1.1' 6379
Multiple slaves
Slaves of slaves

21. Master slave replication
slaveof '192.168.1.1' 6379
Multiple slaves
Slaves of slaves
Scalability

22. Master slave replication
slaveof '192.168.1.1' 6379
Multiple slaves
Slaves of slaves
Scalability
Defer save to slaves

23. Redis Virtual Memory

24. Redis Virtual Memory
Only values swapped out to disk

25. Redis Virtual Memory
Only values swapped out to disk
Most recent values in memory

26. Redis Virtual Memory
Only values swapped out to disk
Most recent values in memory
...disk store

27. Message pub/sub

28. Message pub/sub
Messages are published to channels
PUBLISH chat "Hello Redis!"

29. Message pub/sub
Messages are published to channels
PUBLISH chat "Hello Redis!"
Subscribers declare which channels they’re interested in
SUBSCRIBE chat

30. Message pub/sub
Messages are published to channels
PUBLISH chat "Hello Redis!"
Subscribers declare which channels they’re interested in
SUBSCRIBE chat
Subscribers and publishers don’t care about each other

31. Message pub/sub
Messages are published to channels
PUBLISH chat "Hello Redis!"
Subscribers declare which channels they’re interested in
SUBSCRIBE chat
Subscribers and publishers don’t care about each other
Supports pattern subscription
PSUBSCRIBE f*

32. There are only two hard
things in computer science...

33. DB + caching

34. DB + caching
Multiple copies of data

35. DB + caching
Multiple copies of data
Mainly scaling reads

36. DB + caching
Multiple copies of data
Mainly scaling reads
No higher level functions

37. Redis

38. How we use redis

39. How we use redis
Statistics & Caching

40. How we use redis
Statistics & Caching
Queueing

41. How we use redis
Statistics & Caching
Queueing
Throttling

42. How we use redis
Statistics & Caching
Queueing
Throttling
Inter process communication*

43. When to consider redis

44. When to consider redis
You like it fast

45. When to consider redis
You like it fast
You need more than just key value pairs

46. When to consider redis
You like it fast
You need more than just key value pairs
Small dataset that can fit into memory

47. When to consider redis
You like it fast
You need more than just key value pairs
Small dataset that can fit into memory
Your dataset isn’t critical*

48. • C • Java
• C# • Lua
• C++ • Node.js
• Clojure • Objective-c
• Common List • Perl
• Erlang • PHP
• Go • Python
• Haskell • Scala
• haXe • Smalltalk
• Io • Tcl

50. Thanks!