Typically, the only thing you worry about when you implement and deploy a database is the schema and layout of your data. With CockroachDB this is a little different as the distributed nature of the database allows you to think about resilience and latency at the row level of each table Cockroach delivers bulletproof resilience but also allows you to configure this at the table level.
Further, we can allow you tie data to any node at the row level so you can either meet privacy laws or more importantly meet customer SLAs for transactional latency. How you set the database up will impact both of these and we offer a few different topologies to help you think through these challenges.
In this episode of the Cockroach Hour, we will talk about topologies and the controls you have at the row level to optimize for survivability and latency in CockroachDB.
This episode covers:
- Ranges, replication and the importance of primary keys
- Surviving a data center (region) failure
- Applying a policy to different tables based on row level needs for latency
- Adding survival zones to a production database
- Changing schema in a production instance
- Geo-partitioning and data privacy