Support Models
PostgreSQL Support in South Africa: Running Postgres in Production Without the Guesswork
PostgreSQL is free to download and expensive to neglect. What production-grade PostgreSQL support looks like, the problems that come up most often, and how to decide what your business needs.
By RadixTrie3 September 20263 min read
PostgreSQL has become the default choice for new systems in many South African organisations, and it is increasingly the landing point for migrations away from commercial databases. It is a superb database. It is also one that quietly punishes neglect. This article describes what production-grade PostgreSQL support involves, the problems we encounter most often, and how to decide how much support your environment actually needs.
Why Postgres estates drift
Most PostgreSQL databases start life inside a development team. They are created for one application, configured with defaults, and work well for a long time. Then the data grows, more applications arrive, the people who set things up move on, and nobody is formally responsible for the database tier. The symptoms are familiar: tables that keep growing after data is deleted, queries that were fast last year, disks filling faster than planned, and an upgrade that never quite makes it onto the roadmap.
The problems we see most often
The pattern repeats across industries. These are the situations we are asked to help with most, each linked to a case study describing how we approached it:
<strong>Bloat and runaway storage.</strong> Autovacuum settings that never kept up with the workload, until a database is many times larger than its live data. <a href="/blog/how-radixtrie-helped-a-client-reduce-a-30tb-postgres-database-to-13tb">How we helped a client reduce a 30TB Postgres database to 13TB</a>.
<strong>Sprawl.</strong> Databases nobody can confidently say are still in use, each costing storage, backup time and attention. <a href="/blog/how-radixtrie-helped-a-client-decommission-unused-postgres-databases">Decommissioning unused Postgres databases</a>.
<strong>Unstable performance.</strong> Missing or misused indexes, configuration left at defaults, and workloads that changed after the tuning stopped. <a href="/blog/how-radixtrie-helped-stabilise-postgresql-performance">Stabilising PostgreSQL performance</a>.
<strong>Slow capacity decisions.</strong> Right-sizing approvals that take days because nobody has trustworthy data. <a href="/blog/how-radixtrie-turned-right-sizing-approvals-from-days-into-minutes">Turning right-sizing approvals from days into minutes</a>.
<strong>The skills gap.</strong> Teams with deep Oracle or SQL Server experience discovering that Postgres rewards different habits. <a href="/blog/reskilling-to-postgresql-how-to-navigate-the-challenges">Reskilling to PostgreSQL</a>.
What production-grade PostgreSQL support includes
- Monitoring that someone reads. Connections, replication lag, long-running transactions, bloat, lock waits and disk growth, watched by a person who knows what normal looks like for your systems.
- Backups you can restore. Base backups plus write-ahead log archiving for point-in-time recovery, and restore drills with a measured recovery time. A backup that has never been restored is a hope, not a plan.
- Vacuum and bloat management. Autovacuum tuned per table where the workload demands it, and bloat measured rather than guessed.
- Upgrade planning. The PostgreSQL community supports each major version for roughly five years; PostgreSQL 13, for example, reached end of life in November 2025. Staying on a supported major version is part of running Postgres properly, and major upgrades need testing and a rollback path.
- High availability. Streaming replication with tested failover, so that a server failure is an event and not an outage.
- Security. Role design, authentication configuration, encryption in transit, and auditing appropriate to the data you hold.
- Performance tuning as routine work. Query plans, indexing strategy, memory and checkpoint configuration reviewed continuously rather than after complaints.
Cloud, on-premises or both
Managed cloud PostgreSQL services remove server administration, and that is genuinely valuable. They do not remove the need for database judgement. Parameter tuning, schema and index design, query performance, capacity planning and cost control remain your responsibility, and they are exactly where a senior DBA earns their keep. The same applies in reverse: an on-premises estate benefits from the automation habits that cloud teams take for granted. Good support works across both.
A useful first step is a PostgreSQL health check: versions and support status, backup and recovery posture, bloat, configuration against workload, and a prioritised list of what to fix first.
How much support do you need?
It depends on how much the business depends on the database. A single internal application with a forgiving recovery window needs monitoring, tested backups and periodic reviews. A customer-facing platform needs 24/7 coverage, tested high availability and continuous performance attention. The honest way to decide is to write down, for each database, how long you could be without it and how much data you could afford to lose, then match the support model to those two numbers.
How RadixTrie supports PostgreSQL
RadixTrie is a South African database specialist with senior engineers across PostgreSQL, Microsoft SQL Server, Oracle and MySQL/MariaDB, with offices in Randpark Ridge and Stellenbosch. The model is direct access to senior database experts, no service desk and no escalation loops. For PostgreSQL that covers 24/7 support, managed DBA services, performance tuning, health checks, upgrades and migrations, high availability and disaster recovery, and security and monitoring. Details are on our <a href="/services">services page</a>, and the case studies linked above show the work in practice.
<strong>Related reading:</strong> <a href="/blog/oracle-to-postgresql-migration">Oracle to PostgreSQL Migration: What Actually Takes the Time</a> · <a href="/blog/database-performance-tuning">Database Performance Tuning as a Habit, Not a Project</a> · <a href="/blog/managed-dba-services-explained">Managed DBA Services Explained: What You Hand Over, What You Keep</a> · <a href="/blog/high-availability-disaster-recovery">High Availability and Disaster Recovery: Tested Beats Configured</a>
TAGSPostgreSQLDatabase SupportManaged DBASouth Africa