Perspectives on NoSQL

PGCon 2010

Gavin M. Roy <gmr@myyearbook.com>

Thursday, May 20, 2010

 What is NoSQL?

“NoSQL is a movement promoting a loosely defined class of non-

relational data stores that break with a long history of relational
databases. These data stores may not require fixed table schemas,
usually avoid join operations and typically scale horizontally.
Academics and papers typically refer to these databases as
structured storage.” - Wikipedia

http://en.wikipedia.org/wiki/NoSQL

Thursday, May 20, 2010

 NoSQL: What does it mean?

✤ Coined as the name of a database project in 1998 by Carlo Strozzi

✤ Re-introduced as a general term in 2009 by Eric Evans at Rackspace

✤ Better phrased at No Relationships or No ACID?

✤ Departing from traditional RDBMS technology to scale large traffic

websites

Thursday, May 20, 2010

 What’s new about NoSQL?

✤ Key/Value data stores are not new

✤ DBM (1979), BerkelyDB (1986)

✤ Document Databases are not new

✤ Lotus Notes (1989)

✤ The use case is new...

Thursday, May 20, 2010

 Oxymoron
Scaling large database driven websites is hard

Thursday, May 20, 2010

 Impetus for NoSQL Databases

+
memcached RDBMS

Thursday, May 20, 2010

 Commodity Hardware

“I know of one company that’s managing

to scale portions of their PostgreSQL
servers by purchasing $250,000 servers.
This would cover my 50 node EC2 cluster
for two years.”

- Joe Stump, SimpleGeo

Thursday, May 20, 2010

 Brewer’s CAP Theorem: Pick Two

Consistency

Availability Partition Tolerance

Thursday, May 20, 2010

 “No set of failures less than total network
failure is allowed to cause the system to
respond incorrectly”
- Seth Gilbert & Nancy Lynch, 2002

Thursday, May 20, 2010

 CAP Theorem and PostgreSQL

Consistency

Availability Partition Tolerance

Thursday, May 20, 2010

 CAP Theorem and NoSQL

Consistency

Availability Partition Tolerance

Thursday, May 20, 2010

 fsync = off
a.k.a. Database Administrators running with scissors

Thursday, May 20, 2010

Thursday, May 20, 2010
 Apache Cassandra

✤ Developed initially at Facebook, used at Digg and others

✤ Column-oriented key/value store

✤ Durability via Commit Log, similar to WAL

✤ Eventually Consistent across nodes

✤ Focuses on Availability and Partition Tolerance

✤ Talks via the Thrift protocol, query via Map-Reduce

Thursday, May 20, 2010

Thursday, May 20, 2010
 Apache CouchDB

✤ Major deployments include the BBC and Engine Yard

✤ Key/Value Document Store

✤ Durability by append-only use

✤ MVCC

✤ Focuses on Availability and Partition Tolerance

✤ Talks JSON via HTTP REST

✤ Query by document or JavaScript functions (Map-Reduce)

Thursday, May 20, 2010

Thursday, May 20, 2010
 MongoDB

✤ Major deployments include Sourceforge, Foursquare, Bit.ly, & Github

✤ Key/Value Document Store

✤ No durability without replication

✤ In place updates

✤ Focuses on Consistency and Partition Tolerance

✤ Data stored in BSON (Binary JSON)

✤ Own communication protocol

Thursday, May 20, 2010

 MongoDB & Ad-Hoc Queries

SELECT * FROM things WHERE j != 3 AND k > 10;

db.things.find({j: {$ne: 3}, k: {$gt: 10}});

Thursday, May 20, 2010

 MongoDB & Ad-Hoc Queries

SELECT * FROM things OFFSET 10 LIMIT 10;

db.things.find().skip(10).limit(10);

Thursday, May 20, 2010

Thursday, May 20, 2010
 Project Voldemort

✤ Developed at LinkedIn

✤ Key/Value Document Store

✤ Durability at pluggable data storage layer (BerkeleyDB)

✤ Focuses on Availability and Partition Tolerance

✤ Multiple serialization formats for data

✤ Own communication protocol

Thursday, May 20, 2010

Thursday, May 20, 2010
 Redis

✤ Development sponsored by VMWare

✤ Key/Value Document Store

✤ In-memory database with durability via snapshots

✤ Focuses on Availability and Partition Tolerance

✤ No set data serialization format

✤ Own communication protocol similar to POP3

Thursday, May 20, 2010

Thursday, May 20, 2010
 Tokyo Cabinet/Tyrant

✤ Key / Value DBM Implementation and Network Daemon

✤ Durability via WAL and Shadow Paging

✤ Focuses on Availability and Partition Tolerance

✤ No set data serialization format

✤ Protocols: Tokyo Tyrant Binary Protocol, memcached Compatible Text

Protocol, HTTP REST

Thursday, May 20, 2010

 NoSQL Solves Scaling Problems...

✤ NoSQL users use memcached too!

✤ Reddit uses memcached in front of Cassandra

http://blog.reddit.com/2010/05/reddits-may-2010-state-of-servers.html

✤ Twitter uses memcached in front of Cassandra

http://nosql.mypopescu.com/post/407159447/cassandra-twitter-an-interview-with-ryan-king

✤ Facebook uses memcached in front of Cassandra

http://www.facebook.com/note.php?note_id=39391378919#!/notes.php?id=9445547199

✤ BusinessInsider.com uses memcached in front of MongoDB

http://journal.uggedal.com/tags/mongodb

Thursday, May 20, 2010

 YeSQL
How does PostgreSQL stack up?

Thursday, May 20, 2010

 PostgreSQL Scales

✤ Hot Standby + WAL Streaming adds native read only active standby

✤ Horizontal scaling via sharding using plProxy

✤ Scales with hardware

✤ Projects like pgPool-II and GridSQL

✤ New projects like PostgreSQL-XC

✤ Connection pooling (pgBouncer, pgPool-II)

✤ Replication: Londiste, Slony, Bucardo, Mammoth Replicator

Thursday, May 20, 2010

 KVPBench

✤ http://github.com/gmr/kvpbench

✤ Python app

✤ Benchmarked in OS X 10.6.3

✤ 2.8 GHz Intel Core i7

✤ 8GB DDR3 Ram

✤ 1 Seagate 7200 RPM 1TB Drive

Thursday, May 20, 2010

 KVPBench Details

✤ Supreme Court Decision Data

✤ Default Configurations

✤ Load Test: 100% write, individual inserts, 98k rows

✤ Random Work: 75% Read, 25% Read+Write (Update)

✤ 10 Backends

✤ 10k Queries per Backend

Thursday, May 20, 2010

 Databases Tested

✤ Cassandra 0.6.1

✤ CouchDB 0.11.0

✤ MongoDB 1.4.2

✤ PostgreSQL 9.0b1

✤ Project Voldemort 0.80.2

✤ Redis 1.2.6

✤ Tokyo Tyrant 1.4.44

Thursday, May 20, 2010

 KVPBench Load Times

400.00000

300.00000

200.00000

100.00000

Cassandra CouchDB MongoDB PostgreSQL KV w fSync PostgreSQL KV w/o fSync PostgreSQL w fSync
PostgreSQL w/o fSync Redis Tokyo Tyrant Voldemort

Thursday, May 20, 2010

 KVPBench Random Workload
10k Rows, 10 Workers, Random Workload

150.00000

112.50000

75.00000

37.50000

Cassandra CouchDB MongoDB PostgreSQL KV w fSync PostgreSQL KV w/o fSync PostgreSQL w fSync
PostgreSQL w/o fSync Redis Tokyo Tyrant Voldemort

Thursday, May 20, 2010

 YeSQL?

✤ Amazing community of talented developers who know the subject

matter deeply

✤ Mature code-base that is flexible enough to be a stable platform for

innovation

✤ Fast, stable and well documented code base

✤ Performs as well as (and outperforms some) NoSQL databases while

retaining deeper flexibility

Thursday, May 20, 2010

 Questions?
Follow me on Twitter:
http://twitter.com/crad
Rate this talk:
http://bit.ly/cF04fX
Email me:
gmr@myyearbook.com

Thursday, May 20, 2010