Today we feature a guest blog by Rohit Saboo, Machine Learning Engineering Lead at Nauto. Rohit has spoken at multiple Scylla Summits, and in 2018 his work at Nauto was awarded the Scylla User Award for More Interesting Technical Use Case. How we build trips consistently in a distributed system Early on, I noticed that one of our services would often have errors in retrieving data from the database. Upon investigation, we discovered that some of the queries were being run with full consistency, which in this case was the equivalent of enforcing that if any single node in a […]
We are now introducing a new shard-aware Kafka sink connector for Scylla. This will allow you to use Apache Kafka and the Confluent platform to their fullest advantages.
The remedy to these pain points was Scylla. “Just moving one use case [cart mutations] to Scylla this year we got a huge benefit out of it.” They cut the cluster size from 55 Cassandra nodes to 12 nodes of Scylla, dramatically reducing their EC2 bill.