CloudOps for your databases: a new paradigm for managing your data stores

We went from running databases in a typical LAMP stack, to building databases to run in cloud like environments, to actually just running database in the hosted cloud (or a self-run hosted platform like Kubernetes). The principles of running your data stores, however, have not changed, and still require: monitoring, instrumentation, management, compliance, high availability, optimised performance, security, capacity planning, and keeping up with the SLAs. 

With the advent of being cloud native, maybe in the hosted cloud or via Kubernetes, with a bit of DevOps thrown in, how does doing it the CloudOps way, help change the task of the Database Administrator (DBA)?

We'll cover the MySQL family (MySQL, MariaDB, Percona Server) and PostgreSQL, the Amazon RDS family (RDS for MySQL / PostgreSQL, but also, Aurora), Google Cloud SQL, Oracle Cloud, Azure SQL, Vitess (PlaneScale), and roll your own Kubernetes based solutions. There will also be a smidge of MongoDB (and managing other kinds of data stores). We'll focus on:

* different backup strategies (or do you even need one?)

* planning for multiple data centres (or availability zones) for High Availability

* where the application is hosted

* do you need proxies

* how do you get the most performance?

* what it all should cost?

* monitoring

* security

* cross-cloud databases (aka, hybrid cloud)

* is ML/AI going to change, aka, automate this?

Come to this talk to learn how the DBA task changes, or how Dev can also be closer to Ops when it comes to data stores.