Live Cohort · Sep 21 – Oct 16, 2026

PostgreSQL DBA
Workshop

AlmaLinux 9 administration, high availability, and disaster recovery — taught on a real 3-node PostgreSQL cluster, with SELinux and firewalld left on the whole way through.

36 hours, 14 sections Live + hands-on labs PostgreSQL 18 · AlmaLinux 9
atlas_freight clusterstreaming replication
pg-node1 PRIMARY · WAL sender pg-node2 REPLICA · sync pg-node3 REPLICA · async
getenforceEnforcing
36h
Lecture + lab time
14
Sections, one running case study
3
Node cluster you build & keep
1
Timed incident capstone
Course overview

This is a hands-on, AlmaLinux 9-first administration course for anyone who already knows their way around a Linux shell and now needs production-grade PostgreSQL depth: installation and configuration, backup and recovery, performance tuning, and — further than most "fundamentals" courses go — real streaming replication and Patroni-based high availability, disaster recovery, and automation.

This course also keeps SELinux enforcing and firewalld active on every node, start to finish. Plenty of tutorials tell you to disable both to "get PostgreSQL working." This one shows the real fix each time — semanage fcontext, restorecon, audit2allow, firewalld rich rules — so what you build here survives contact with an actual RHEL-shop security policy.

Running case study

Atlas Freight Systems

A mid-size logistics company migrating its shipment-tracking platform onto a 3-node PostgreSQL cluster. Labs compound — the cluster you provision in Section 0 is the one you're still operating in the Section 13 capstone. Nothing resets between modules, because that's how a real production system behaves.

Who this course is for

  • Linux sysadmins or generalist DBAs comfortable at the command line but new to PostgreSQL specifically.
  • Developers moving into a DBA or platform-engineering role who need real operational depth, not just SQL syntax.
  • Anyone preparing for a production PostgreSQL on-call rotation on RHEL-family infrastructure.

Prerequisites

  • Comfortable with the Linux command line: filesystem, editing files, services, basic networking.
  • Basic SQL (SELECT / INSERT / UPDATE / DELETE) — this course teaches administration, not introductory SQL.
  • No prior PostgreSQL, AlmaLinux, or SELinux experience required — those are taught in context as you go.
By the end

What you'll be able to do

Install, configure, and secure PostgreSQL 18 on AlmaLinux 9 from a bare VM, the DBA-standard way — PGDG repository, hardened pg_hba.conf, SCRAM auth.
Design a least-privilege role hierarchy and implement Row-Level Security.
Configure WAL archiving and perform a real point-in-time recovery, plus run pgBackRest for ongoing backups.
Diagnose and fix table bloat, slow queries, and I/O bottlenecks using real statistics, not guesswork.
Build streaming replication by hand, then automate failover with Patroni and etcd — and prove it under a simulated outage, with SELinux still enforcing.
Perform logical replication and in-place major version upgrades, and write DR runbooks and incident RCAs a team could actually use.
Program format

Live sessions, plus labs on your own cluster

Two live sessions a week walk through the concepts and demo the trickiest steps; the hands-on labs are worked independently, against the same 3-node AlmaLinux cluster, between sessions.

Dates
Sep 21 – Oct 16, 2026
Cadence
4 weeks · 2 live sessions/wk
Schedule
Tue & Thu, 6:30–8:30 PM EAT
Delivery
Live online + recordings
Curriculum · 36 hours across 14 sections

Four weeks, one continuous cluster

Nothing resets between weeks — the environment you provision in Week 1 is the same one you're diagnosing under pressure in the Week 4 capstone.

Environment & tooling

What this course actually runs on

ComponentThis course uses
Operating systemAlmaLinux 9 — a free, 1:1 RHEL-compatible rebuild; every command uses dnf, systemd, and firewalld.
PostgreSQL versionPostgreSQL 18 via the PGDG yum repository; notes flag anything specific to 16/17 for learners on older production clusters.
Security postureSELinux enforcing and firewalld active on every node for the entire course — labs teach the real fix, never setenforce 0.
Lab topology3 AlmaLinux 9 VMs or cloud instances (pg-node1/2/3), 2 vCPU / 4 GB RAM / 20 GB disk minimum each.
Connection poolingPgBouncer
Backup toolingpg_dump / pg_basebackup (native) and pgBackRest (production-standard).
High availabilityNative streaming replication, then Patroni 4.x + etcd for automated failover.
AutomationBash, systemd timers, Ansible (dnf module).

Local VMs (VirtualBox + Vagrant, the almalinux/9 box) or the smallest AlmaLinux 9 image from any cloud provider both work — the lab environment is on your own machine or account, not provided as part of the course.

Your instructor

Taught live, not pre-recorded

SO

Steve Osine, CDMP

Trion Consulting Group

Steve is the Founder and Principal Trainer at Trion Consulting Group, with 14+ years of hands-on experience deploying PostgreSQL in production — including high-availability clusters, disaster recovery, and migrations. He holds the CDMP (Certified Data Management Professional) credential and is a Certified DBA across multiple data platforms, including Oracle, PostgreSQL, and SQL Server. He teaches the way he works in the field: SELinux and firewalld stay enforcing throughout, because that's what actually survives a real production audit.

Enrollment

Reserve your seat

Prices in Kenyan Shillings. Pay securely by M-Pesa or card, powered by IntaSend.

1Choose a plan
2Your details & payment
Total due today KES 29,000
Enter your first name
Enter your last name
Enter a valid email address
Enter the phone number to pay from
Payments are processed by IntaSend; card and M-Pesa details never touch this page directly. Questions before you pay — reach us below.
Questions

Before you enroll

KES 29,000early bird, ends Sep 14
Enroll now