Case Study: Mobilizing Marketo Engineering to transform the Multi-tenant SaaS Campaign Processing Engine to a Reactive System that Achieves Excellent Scalability and Throughput

Oct 24, 2018

Speakers

About

Marketo customers have grown bigger and bigger and are demanding exceptional marketing campaign processing power, given finite compute resources. Our legacy campaign processing engine, built on top of LAMP, is not going to cut it. While Marketo has transformed some legacy systems using Big Data technologies like Spark and HBase, these technologies are not a good fit for our next generation campaign processing engine. We decided to embrace Reactive but the journey was not without challenges: - First, we needed to complete multiple prototypes to prove to the architects that Reactive system was a better fit. - Second, we had little prior knowledge in reactive programming. We put together a training plan on various topics, including Scala, Akka cluster, Reactive Programming methodologies, etc. - Third, it is important for our new campaign processing engine to co-exist with the legacy one so we can migrate the sub-components one by one, while not interrupting services for customers. In the end, it has been a success and many Marketo customers have been migrated to the new campaign processing engine and are enjoying the spectacular scalability and throughput.

Organizer

Categories

About Lightbend

The Lightbend Reactive Platform is a JVM-based runtime and toolset for building Reactive Applications.

Store presentation

Should this presentation be stored for 1000 years?

How do we store presentations

Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%

Sharing

Recommended Videos

Presentations on similar topic, category or speaker

Interested in talks like this? Follow Lightbend