This document provides guidance on how to successfully implement Apache Cassandra in a production environment without issues. It recommends starting with a small, well-defined project like monitoring website events or users, rather than trying to build a large, multi-year platform. The document outlines choosing a specific pain point to address, implementing a simple proof of concept using Cassandra for tasks like event tracking, and iterating from there. It cautions against copying relational data models into Cassandra and emphasizes understanding how Cassandra works differently from SQL databases. The goal is to start small and grow capability over time rather than taking on too much at once.

1. Patrick McFadin, Chief Evangelist for Apache Cassandra
@PatrickMcFadin
How to roll Cassandra into
production without losing your
health, mind or job

2. 2© 2015. All Rights Reserved.
Help!

3. 3© 2015. All Rights Reserved.
Strategy Stop believing
in unicorns
Reaching Maturity

4. Strategy

5. Scaling everest?
5© 2015. All Rights Reserved.
You are here
You want to be here

6. One step at a time
6© 2015. All Rights Reserved.

7. 7© 2015. All Rights Reserved.
Nope!

8. 8© 2015. All Rights Reserved.

9. The problem (too much)
9© 2015. All Rights Reserved.
“We’re building a platform”
“It’s a 3 year project”
“We’ll build it from scratch with
all the lessons learned from the
last app”
“Rewrite everything using technology X”

10. 10© 2015. All Rights Reserved.
Fail

11. Pick an app
11© 2015. All Rights Reserved.
User
Administration
Product
Web Site
Fraud
detection
Website and
user monitoring

12. Pick a pain point
12© 2015. All Rights Reserved.
Website and
user monitoring
Event tracking Alerting Log storage

13. Pick an implementation
• Kafka
• Cassandra
• Spark
13© 2015. All Rights Reserved.
Event tracking

14. Stick the landing
14© 2015. All Rights Reserved.
Yay a new project!
Not sure about this!
Done!

15. Move on to the next thing
15© 2015. All Rights Reserved.
Event tracking Alerting

16. Stop believing in unicorns

17. Cassandra is not…
17© 2015. All Rights Reserved.
A Data Ocean or Pond., Lake
An In-Memory Database
A Key-Value Store
A magical database unicorn that farts rainbows

18. 18© 2015. All Rights Reserved.
When to use…
Loose data model (joins, sub-selects)
Absolute consistency (aka gotta have ACID)
No need to use anything else
You’ll miss the long, candle lit dinners with your Oracle rep
that always end with “what’s your budget look like this
year?”
Oracle, MySQL, Postgres or <RDBMS>

19. 19© 2015. All Rights Reserved.
When to use…
Uptime is a top priority
Unpredictable or high scaling requirements
Workload is transactional
Willing to put the time or effort into
understanding how Cassandra works and
how to use it.
Use Oracle when you want to count your money.
Use Cassandra when you want to make money.
Cassandra

20. Copy n Paste your relational model
20© 2015. All Rights Reserved.
APACHE
CASSANDRA

21. 8 Fallacies of distributed computing
21© 2015. All Rights Reserved.
1. The network is reliable.
2. Latency is zero.
3. Bandwidth is infinite.
4. The network is secure.
5. Topology doesn't change.
6. There is one administrator.
7. Transport cost is zero.
8. The network is homogeneous.

22. 22© 2015. All Rights Reserved.
Test Test
Test
Test
Test
Test
Test
Test
Test
Test
Test
TestTest
Test
TestTest
Test

23. 23© 2015. All Rights Reserved.
APACHE
CASSANDRA
API Layer
Cache
Load
Balancer
Load Test

24. 24© 2015. All Rights Reserved.
APACHE
CASSANDRA
API Layer
Cache
Load
Balancer
Load Test

25. Maturity

26. The moment!
26© 2015. All Rights Reserved.

27. Other teams come to see you
27© 2015. All Rights Reserved.

28. Center of excellence
• Group dedicated to providing
leadership and guidance
28© 2015. All Rights Reserved.

29. 29© 2015. All Rights Reserved.

30. Thank you!@patrickmcfadin