This document summarizes a New York Redis Meetup event. It introduces Aleksandr Yampolskiy and Danny Gershman, who will discuss Redis, a key-value store that can be used for caching, publishing/subscribing, and as a data store. Redis allows for fast, in-memory storage of data structures like strings, hashes, lists, sets and sorted sets. The document provides an overview of Redis' capabilities and common uses, such as caching, real-time analytics, and AOP caching. It also notes that Cinchcast is hiring for backend architect and frontend engineer roles.

1. New York Redis Meetup
Welcome Session
June 26, 2012
@ayampolskiy
@dannygnj

2. Who Are We?
Aleksandr Yampolskiy
•CTO of Cinchcast, BlogTalkRadio, and Cinch.FM
companies that provide solutions to create, share, measure,
and monetize audio content.
•Previously head of security and compliance for Gilt
Groupe companies
•Various leadership roles in Goldman Sachs, Oracle,
Microsoft building scalable, enterprise software for IDM,
SSO, AuthN/AuthZ.
•Ph.D. in Cryptography
Danny Gershman
•Senior Software Engineer at Cinchcast
•Previously held position as Lead Software Engineer at The
NewsMarket and Synaptic Digital
•Systems Architect at Innovative Systems & Solutions
•Specializing in web technologies, caching, and scale.

3. What is This Meetup?
• New York REDIS NOSQL Meetup
• Topics of interest: REDIS, NOSQL,
Engineering, Entrepreneurs, Open-Source

4. What Do You Want to Get Out Of It?

5. What is REDIS?
• Database?
• Caching?
• Pub/Sub?
• Key/Value Store?
• Redis is more like a flexible
tool designed to solve
specific problems

6. REDIS
REmote DIctionary Server (www.redis.io)
Key-Value Store “on steroids”
Written by Salvatore Sanfilippo (@antirez) in
2009. Now up to 2.4.13

7. Redis Manifesto
– Redis is a DSL (Domain Specific Language) that manipulates
abstract data types and implemented as a TCP daemon. Keys
are binary-safe strings and values are different kinds of
abstract data types.
– Redis has persistence option but Memory storage is #1.
– The Redis API is a direct consequence of fundamental data
structures.
– Code is like a poem.
– We believe designing systems is a fight against complexity.
Most of the time the best way to fight complexity is by not
creating it at all.
– Redis API has two levels: 1) a subset of the API fits naturally
into a distributed version of Redis and 2) a more complex API
that supports multi-key operations.
– We optimize for joy. When there is no longer joy in writing
code, the best thing to do is stop.
7

8. REDIS is Fast
• Fast ~100,000 queries per sec.
• Compare it with ~30,000 queries per sec
for MS SQL Server (if you are lucky)
• Kind of like memcached, but values can
be persisted and all keys are kept IN
MEMORY.

9. Modern Computer Latency
L1 3 cycles
Non-blocking
L2 14 cycles
RAM 250 cycles
DISK 41,000,000 cycles
Blocking
NETWO 240,000,000 cycles
RK
Ryan Dahl, Yahoo BayJax meetup, May 2010

10. Redis VS Memcached
10

11. Who is using REDIS?
11

12. Data structures
Keys are strings which identify pieces of data
(values)
Values can be strings, integers, hashes, lists,
sets, or sorted sets
Various atomic operations (push, pop, add,
read, set, etc.) – after all REDIS is single-
threaded
12

13. Many Language Bindings

14. REDIS is Simple

15. Scale
Horizontal (multiple processes)
Vertical (master-slave)
Master-master via zookeeper

16. Master-Master Replication
Native Redis Master-Master replication doesn’t
exist yet, slated for release end of year
Solution using Apache Zookeeper (
http://zookeeper.apache.org/)
Checks for status using BLPOP
Will promote a slave to master

17. When Not to Use Redis
Data is larger than memory
Need ACID requirements
Big data

18. Security Concerns
It is designed to be used only inside trusted
environment (kind of like MongoDB)
Basic password authentication (but it’s sent in
cleartext)
No data encryption support
18

19. Security Recommendations
Configure firewall to block inbound access on
REDIS ports
Don’t run REDIS server as root, and harden the
server it’s on
Don’t store sensitive data in there.
19

20. AOP Caching
• Dynamically construct a hash key from argument
values.
• “Round up” volatile keys such as timestamps
• Put into REDIS cache with a configurable TTL
• Lookup from cache before SQL server + lazy writes.

21. Real-time Analytics
Communicate from hydrating application
NodeJS listens to Redis and transmits data
to a page using Socket.IO
Historical data is stored in Redis
Pub/Sub is used to notify listening sockets

22. Studio Callers

23. Cinchcast is Hiring
• Well-funded profitable startup Open Positions:
used by millions.
•Back-end Architect
• Patented technology utilized in a
new way. •Lead Front-end Engineer
• Medical/dental/etc. benefits.
• Great office space in mid-town
•If you are good we’ll find a role
right near subway.
• Flexible hours. Top-notch
for you : jobs@cinchcast.com
compensation + stock options.