Apache Cassandra is a highly scalable, resilient, distributed database. This has been the backbone of multiple companies, like Netflix, Apple, or Grubhub, where it’s crucial to have quick writes, with a guarantee that the operation will be given the best opportunity to complete; also, it can be a nightmare to administer as latency, network settings, disk management can have a huge performance impact. In this talk Carlos will share experiences and lessons learned while applying Google’s SRE principles and practices in the administration of hundreds of hosts used for multiple Cassandra clusters