mParticle and ScyllaDB attended The NoSQL & NewSQL Database Meetup at the AWS Loft in NYC. Yuan Ren from mParticle gave a nice presentation that explained their journey from Apache Cassandra to Scylla and how they process 50 billion monthly messages with full availability and performance.
mParticle provides a single, secure API to integrate and orchestrate an organization’s entire marketing stack so that brands can enhance analytics and optimize acquisition, engagement, and monetization in a multi-screen world.
The mParticle Platform needed a way to process a lot of information very quickly and began with Apache Cassandra to handle their data loads. To better understand their workload, mParticle currently has over 1 Billion unique smartphone users per month. The monthly data volume consists of 50 billion batches, 100 billion events, and 150TB of data in binary format added to Amazon S3.
With Apache Cassandra, mParticle got high read and write throughput, commercial support, horizontal scalability, and a product with a proven track record. However, they ran into problems along the way. Running Apache Cassandra required many human resources to maintain and tune it properly, commercial support was expensive and not so great, and they experienced performance and latency issues.
After learning about Scylla, mParticle started a proof-of-concept deployment on the same hardware to see if Scylla could solve their problems. They loved the fact that Scylla is compatible with Apache Cassandra and did not require any code changes to their software and they would no longer have to worry about tweaking the Java Virtual Machine (JVM).
They now have a NoSQL cluster comprised of 10 nodes with a 100 TB data storage and have concluded that Scylla is a better solution for them because…
- Scylla has a much lower compaction backlog
- Configuration and self-tuning capabilities are easier in Scylla
- They are achieving lower latencies and better performance with Scylla – peaking at 6 GB/s compaction throughput per node while providing low latency
- ScyllaDB has a knowledgeable and responsive support team
- They easily migrated without service disruptions and had to make zero client side changes
In particular, mParticle had good things to say about their experience with the support they received from ScyllaDB. “Their engineers know the ins and outs of their product, without having to ‘get back to you’ hours later,” the director of DevOps said. “All engineers are responsive and on top of an issue or question so you get the response, and ultimately the resolution, as fast as it can possibly be done. They are absolutely dedicated and reliable and will make sure that your issue is resolved, or they will work with you 24/7 until you are satisfied.“
To view the presentation and learn more in depth, please check it out on SlideShare or watch the video of the meetup:
Apache®, Apache Cassandra®, are either registered trademarks or trademarks of the Apache Software Foundation in the United States and/or other countries. No endorsement by The Apache Software Foundation is implied by the use of these marks.