Database Management (Postgres, MySQL, Redis)
About This Service
Postgres, MySQL & Redis Database Management for UAE Companies
I manage and tune production databases — PostgreSQL, MySQL/MariaDB, and Redis — for businesses in Dubai, Abu Dhabi, and Sharjah. The most common engagement starts with performance: slow-query analysis using pg_stat_statements or the MySQL slow query log, missing or bloated indexes, and query plans that scan millions of rows where an index lookup would do. A correctly indexed database routinely turns 8-second pages into 80-millisecond pages, which for an e-commerce or booking platform translates directly into kept customers and AED revenue.
Equally important is the unglamorous work most setups skip: automated backups with restores that are actually tested (an untested backup is a hope, not a plan), streaming replication and failover so a single server death is not a business event, connection pooling with PgBouncer so traffic spikes do not exhaust connections, and Redis as a caching layer in front of expensive queries. I also perform major version upgrades — for example Postgres 13 to 16 — using logical replication so the switchover happens without downtime.
Every engagement ends with a data integrity audit: orphaned rows, missing foreign keys, duplicate records, and charset/timezone inconsistencies that quietly corrupt reports. For free-zone and mainland SMEs holding customer data in the UAE, knowing the database is consistent, backed up, and restorable is the baseline for everything else.
What's included
- Performance tuning — Slow-query analysis, index design, and query rewrites with before/after timings
- Backups with tested restores — Automated daily backups plus a documented, rehearsed restore — not just a cron job
- Replication & failover — A streaming replica and a written failover procedure for when the primary dies
- Redis caching layer — Cache-aside setup for your hottest queries with sensible TTLs and invalidation
- Zero-downtime upgrades — Major version upgrades via logical replication, switched over without an outage
- Integrity audit report — Orphans, duplicates, and constraint gaps found and fixed, with a written summary
How it works
- 1Health check
Read-only access first: I profile query load, index usage, backup status, and replication state, and send you a findings report.
- 2Agree the fixes
We prioritise together — usually indexes and backups first, then pooling, caching, and upgrades — so nothing changes without your sign-off.
- 3Apply in a maintenance window
Changes are applied with rollback steps prepared; risky operations run on a replica or staging copy first.
- 4Verify & document
Before/after metrics, a tested restore drill, and a runbook your team can follow at 3am.
Why work with me
| With me | Typical agency | |
|---|---|---|
| Restore actually rehearsed, not assumed | ||
| Before/after query timings in the report | Rarely | |
| Read-only diagnosis before any change | ||
| Senior DBA attention on your data | Direct | Junior on rotation |