Managing PostgreSQL with Puppet
summary
Afilias uses Puppet plus a Git workflow and Hiera to install and configure PostgreSQL at scale, managing over 100 production database servers, replicas via Slony, cron jobs and packaged installs; a custom pgs_cluster resource handles initdb, data directories and user creation because community modules didn’t support multiple minor versions; configurations live in Hiera YAML hierarchy; PG basebackups or S3 restores seed new nodes, and remove classes undo resources for safe teardown.
AI-written summary · about
chapters (30)
- 0:00Managing PostgreSQL with Puppet
- 1:02Afilias
- 2:52100+ of Databases
- 3:47Technologies
- 5:31DevOps
- 6:44Deployment Automation
- 7:27Goals of Automation Database servers 'look the same
- 8:16Consistency between QA, Staging. Production
- 9:15Reduce Manual Commands
- 10:29Puppet Architecture
- 11:12Puppet Terminology
- 12:13Puppet Class
- 14:01Declarative
- 15:06Puppet At Alilias
- 18:49Goals for Databases
- 19:14PostgreSQL Module
- 20:52pgsql cluster resource
- 21:27Templates
- 24:30Hiera 101 Separate Code From Data
- 25:24Hiera 101: Lookup Values
- 25:57Hiera 101: Define a Hierarchy
- 26:52Hiera 101: Hierarchy
- 27:56Hiera 101: postgres.yal
- 29:17con template server type
- 30:51Switch Based Contig
- 35:24Slony Settings
- 36:00Slony: Slonik preambles
- 37:06Removing Things
- 41:46Working with others
- 43:49Questions?
related talks
55mManaging PostgreSQL With Ansible
13mBuilding reliable infrastructure with Postgres
36mPostgreSQL: Ups, DevOps...
46mBig databases, many servers, data center or cloud - it's all here!
36mPostgreSQL: Ups, DevOps...
29mBuilding Enterprise Grade PostgreSQL using Percona distribution for PostgreSQL
55mHigh Availability with PostgreSQL and Pacemaker
46mPostgres Platform "Best Practices" for the Modern DBA