Cassandra Day London 2015

Showing how Cassandra resilience can keep your data save even when a multitude of failures happen. This is a description how our team kept a 15 node cluster running (3 nodes per DC, 5 DCs) when a multiple DC fail happen and the cluster was already stretched out resource wise.

Building a Microservices platform takes time and challenges a number of traditions in our industry. Coming to grasp with the Brewer's theorem, BASE, streaming, and de-normalisation of data is a crucial step in accomplishing a successful Microservices platform.

There's a lot of theory on Microservices out there, and with your nose up to the grindstone and progress slow it can at times be difficult to relate to and to remember how important it all is. With a run through of development styles of streaming, raw event stores, event sourcing, graph databases, and drilling down to infrastructure components like Graphite, Kibana, and Zipkin, this talk will share with you the experience needed to implement Microservices, with Cassandra as a cornerstone technology, in a modern polyglot persistence enterprise platform.

With 6 years experience breaking up a monolith java+sql enterprise dinosaur into a Microservices platform today existing with over 200 thrift and http services – Mick can give a hands on account of facing these challenges.

British Gas Connected Homes (the creators of Hive Active Heating) embarked last year in the creation of a unified platform for all their different IoT projects. The company is a subsidiary of British Gas specialising in connected devices in the home and handles data ranging from smart gas and electricity meters to connected thermostats and boilers. Jim Anning and Josep Casals will go through their experience building a unified platform with Apache Cassandra, Apache Spark and related technologies. Focus will be of the main lessons learnt and the things they wish they had known beforehand.

This will be a classic "use-case" story of how Cassandra fits in to Hailo, how it's enabled us to run our business, and how its made our lives easier.

I'll include a story of how our use of databases has evolved over time.

I will also give some detail on the way we've learned to tackle some common problems:
* Coping with Micro-services
* Automation
* Monitoring of Cassandra (hopefully with an introduction to a newly open-sourced monitoring tool written in-house)
* Backing it up and restoring data

Client: "We want you to build us a Big Data platform, capable of analysing billions of transactions in real time". 

Semblent: "Ummm, ok". 

Client: "We will only pay for three developers...". 

Semblent: "Errr, ok". 

Client: "....and by the way, you have 4 months". 

Semlent: "Owch.... ok".

We did it... and here's how!

This talk will introduce the various options around securing Cassandra and DataStax Enterprise. Attendees will gain an understanding of the various features and options available for protecting systems using Cassandra, OpsCenter and DataStax Enterprise.

You wouldn't try and use a saw to hammer a nail so why do people try to use databases as queues? Or perhaps more subtly: rely on atomic updates to mutable data distributed across the world (when you say it out loud is sounds a bit ridiculous right?)

In this talk we're going to go through the most common ways to misuse Cassandra. Some are obvious and you should be using a different kind of technology, like the queue example, but others you just need to change the way you're using Cassandra.

Topics we'll cover:
* Joins (yes we'll get started with the obvious one)
* Mutable data / read then write / breaking the laws of physics
* Really wide rows (but the Internet said wide rows were good?)
* Batches!
* Multi partition queries a.k.a ReadTimeout heaven
* Secondary indexes under the covers
* Tripping over tombstones

Crazy retry policies: I am looking at you Downgrading retry policy
Misusing technology isn't isolated to the distributed database world, I used to work on a small piece of queueing software called WebSphereMQ and what do you know? A good portion of our customers used it as a database!