Hydra geo-redundant failover for Postgresql
Jose Cores Finotto
Engineering Manager at Groupon for relational databases, leading efforts for development of solutions and integrations between all the Groupon infrastructure ecosystem, principally focused on the database aspect. With more than 15 years of experience in development and database administration, Jose is really interested in automation, and improvement of processes.
Álvaro is a passionate database and software developer. Founder of OnGres, he has been dedicated to PostgreSQL and R&D in databases for more than two decades.
Álvaro is at heart an open source advocate and developer. He has created software like StackGres, a Platform for running Postgres on Kubernetes. As a well-known member of the PostgreSQL Community, Álvaro founded the non-profit Fundación PostgreSQL and the Spanish PostgreSQL User Group. He has contributed, among others, the SCRAM authentication library to the Postgres JDBC driver.
You can find him frequently speaking at PostgreSQL, database, cloud (becoming an AWS Data Hero in 2019), and Java conferences. In the last 10 years, Álvaro has completed more than 100 tech talks (https://aht.es).
No video of the event yet, sorry!
PostgreSQL High Availability (HA) is a very desirable goal for most, if not a strict requirement. And there are technologies and literature to help provide HA on a PostgreSQL cluster.
However at Groupon we manage hundreds of database servers, distributed across several datacenters in the world. And our internal policy mandates that we need a Disaster Recovery mechanism to switch a complete datacenter over to another.
With this main requirement we have built Hydra, a soon-to-be-open-sourced solution that implements geo-redundant failover. It relies on Consul for distributed consistency and as the basis for the multi-datacenter approach, and is implemented as a lightweight agent in Java8.
- 50 min
- PostgresConf US 2018
- Operations and Administration