MyDBA.dev vs. Redgate Monitor · PostgreSQL, head to head · 2026
Deep Postgres, not a bolt-on

Redgate watches your whole estate. MyDBA lives inside Postgres.

They're built for different jobs. If you need to watch SQL Server, Oracle, MySQL and MongoDB alongside Postgres, Redgate Monitor is the right call — one mature tool across your whole estate, in a single pane. That breadth is also the trade-off: Postgres is one engine among five. MyDBA does one thing — it goes deep on PostgreSQL, and only PostgreSQL, so its advice is specific, advanced, and tailored to how Postgres actually behaves: automatic incident post-mortems, baseline anomaly detection, verified query fixes, and first-class extension monitoring. Here's the head-to-head where it counts most — incident analysis, anomaly detection, query tuning and extension monitoring — plus everything else.

Jump to the deep-dives → Where Redgate wins Postgres-only · by design
The five-second scoreboard

Four areas that matter, side by side

Both are capable monitors. The gap opens up exactly where Postgres gets deep — automatic diagnosis, anomaly detection, verified tuning, and the modern extension ecosystem.

🐘 MyDBA
Area
Redgate Monitor
Automatic post-mortems — root cause, timeline, blast radius, evidence-linked, PDF export▸ full incident engine
Incident analysis
Real-time activity + wait charts to investigate by handno automated post-mortem
Baseline anomaly detection that auto-promotes sustained deviations to incidents▸ 8 detectors, deterministic reports
Phantom incidents
Threshold/baseline alerts + wait-event visualizationno auto-anomaly incidents
AI tuning grounded in your metrics, plus a verifier that checks each suggestion▸ + fix-query & SQL fixes
Query tuning
"Analyze Query" AI advice from query text + plansolid, but query-scoped only
First-class TimescaleDB, pgvector, PostGIS, pg_cron, pg_partman monitoring▸ nobody else does this
Extension monitoring
Core Postgres metrics onlyno extension-specific views
PostgreSQL, exclusively — every check, view and recommendation built for PG internals▸ deeper, Postgres-specific advice
Focus
5-engine estate in one pane — SQL Server, Oracle, MySQL, MongoDB + Postgresthe right pick for mixed shops
Deep-dive 1 of 4

Incident analysis

When something breaks, the question is never "what did the CPU graph do" — it's "what happened, why, what did it touch, and what do I do now." This is where the two tools diverge most.

The core difference

An automatic post-mortem vs. a place to investigate

Redgate gives you excellent real-time and historical data to run the investigation yourself. MyDBA runs the investigation for you and hands you the write-up.

🐘 MyDBAAutomated
  • AI Incident Analysis generates a full post-mortem: root cause, an ordered timeline, blast radius, and remediation steps.
  • Every claim deep-links to raw evidence — the exact metric window, queries, and log lines it was drawn from.
  • Multi-window evidence correlated automatically, with PDF / HTML export and full history to hand to your team.
  • Grounded in real collected context via server-side RPCs — not a generic prompt over a screenshot.
Redgate MonitorManual
  • Current Activity shows queries running right now and flags blocking / contention in real time.
  • Wait-event visualization and historical metrics let a DBA reconstruct what happened — by hand.
  • Strong alerting library tells you that something breached, with the metric context around it.
  • No automated root-cause report, no evidence-linked timeline, no exportable post-mortem artifact.
VerdictMyDBA. Redgate gives you the raw material to diagnose; MyDBA delivers the finished diagnosis — the single biggest day-to-day difference at 2am.
Deep-dive 2 of 4

Phantom incidents

Threshold alerts catch only what you thought to configure. They miss slow degradations that never cross a line, and anomalies on metrics nobody wrote a rule for.

Anomaly detection, done properly

"This is weird for this database" — caught automatically

MyDBA's Phantom Incidents learn each database's normal behaviour per metric and per hour-of-day, then promote sustained deviations into real incidents. Redgate stays in the threshold-and-baseline-alert world.

🐘 MyDBAUnique
  • Baseline-relative detection across 8 detectors — latency, lock-waits, connection saturation, error rate, replication lag, deadlocks, temp-file spill, rollback ratio.
  • Sustained-only by design (robust MAD sigma + hour-of-day median) — single-minute blips are ignored, so it isn't noisy.
  • Each promotion gets a deterministic report with zero AI required; AI enrichment is an optional one-click overlay.
  • Recurring patterns collapse into a single meta-signal so you see chronic conditions, not a wall of duplicates.
Redgate MonitorNot offered
  • Threshold and baseline alerts you configure — the classic "warn if X > N" model, with a large built-in alert library.
  • Wait-event charts help you spot anomalies visually — but you have to be looking.
  • No engine that learns per-database normal and auto-raises an incident when behaviour drifts.
  • Slow degradations that never breach a configured threshold can pass unnoticed.
VerdictMyDBA, clearly. This is a category Redgate simply doesn't have — anomaly detection that turns "the database feels off" into a documented incident.
Deep-dive 3 of 4

Query tuning

The fairest fight of the four — Redgate ships real AI query analysis here. But "explain this one query" and "fix it, verified, in context" are different jobs.

Both have AI — the depth differs

Advice on a query vs. a verified fix in context

Redgate's "Analyze Query" is genuinely useful. MyDBA grounds its tuning in your collected metrics, generates the fix, and then independently verifies the suggestion before you trust it.

🐘 MyDBADeeper
  • Metrics-grounded tuning — explain-plan, fix-query and tuning recommendations fed real query + database context, not just the SQL text.
  • A verifier double-checks AI query suggestions and shows you a verdict badge before you act on them.
  • EXPLAIN Plan Advisor with five plan views (tree, flame, waterfall, bottleneck, cost-vs-actual) and plan-regression detection.
  • Query resources broken down by I/O / memory / CPU (via pg_stat_kcache), plus log-based samples with real parameter values.
Redgate MonitorCapable
  • "Analyze Query" AI advice — walks the likely root cause and suggests fixes from the query text plus execution plan.
  • Visual query-plan viewer — reads plans from the log (auto_explain), correlates with pg_stat_statements, shows expensive ops and planned-vs-actual rows.
  • Top-queries ranking by execution count, duration or I/O time — good for finding the expensive ones.
  • Advice is query-scoped and un-verified — no independent check, no fix-in-context workflow, no regression tracking.
VerdictMyDBA on depth — grounded context, a verifier, and regression detection — while giving Redgate genuine credit for shipping solid AI query analysis of its own.
Deep-dive 4 of 4

Extension monitoring

Modern Postgres is its extensions. If you run TimescaleDB, pgvector, or PostGIS, generic Postgres metrics stop telling you what you need to know.

The cleanest gap of all four

Dedicated extension surfaces vs. core metrics only

This is where MyDBA is not just deeper but categorically different — purpose-built views for the extensions people actually build on. Redgate monitors core Postgres and stops there.

🐘 MyDBAFirst-class
  • TimescaleDB — hypertables, chunk-level visibility, compression ratios, continuous aggregates and background jobs.
  • pgvector — IVFFlat / HNSW index inventory, vector query performance, and probes / ef_search audits.
  • PostGIS — spatial inventory, GiST / SP-GiST index health, and seven geometry-validity health checks.
  • pg_cron & pg_partman — scheduler health and partition-management monitoring, with a Setup Assistant that detects and verifies each extension.
Redgate MonitorCore only
  • Solid core Postgres monitoring — vacuum, XID wraparound, database growth, connections, waits.
  • Uses standard extensions (pg_stat_statements, auto_explain) as data sources — not as monitored subjects.
  • No dedicated views for TimescaleDB, pgvector or PostGIS — a hypertable or an HNSW index looks like any other table.
  • Time-series, vector-search and geospatial workloads are effectively invisible at the extension level.
VerdictMyDBA, no contest. If your Postgres runs modern extensions, this alone is the deciding factor.
Everything else

The rest of the picture

Beyond the four headline areas, here's how the broader Postgres feature set compares. "Partial" means it exists but in a narrower or secondary form.

First-class ~ Partial / add-on Not offered
Capability🐘 MyDBARedgate Monitor
Automated health checks with SQL fixes75+ checks, 10 domains, P1–P4 ranked copy-paste fixes ~built-in advice
Index advisorunused · missing · duplicate, cluster-aware replica-safe ~unused-index flags
Cluster topology + logical replicationauto-discovery, PG17 conflict detection 10 repl. checks ~replication metrics
Provider-aware recommendationsRDS · Aurora · Neon · Supabase · Cloud SQL… 12 providers
Real-time activity & blocking view Current Activity
XID wraparound monitoringwith blocker detection + recovery SQL + auto recovery SQL ~alert + visualization
Smart alertingrate-of-change + compound conditions sparkline per event large alert library
Multi-engine estateSQL Server · Oracle · MySQL · MongoDB Postgres only Redgate's strength
Free-forever tierone primary, all non-AI features, no credit card trial / paid
Enterprise maturity & track recordyears in production, large install base ~newer since 1999
Playing fair

Where Redgate is the better choice

A comparison you can't trust is worthless. Redgate is a serious, mature product — here's where it genuinely wins, and where you should pick it over MyDBA.

Multi-database estates in one place

This is Redgate's decisive win. If you run SQL Server, Oracle, MySQL or MongoDB alongside Postgres, Redgate monitors every engine from a single pane of glass — one tool, one login, one alerting model across the whole estate. MyDBA is Postgres-only by design and cannot be that single tool for a mixed shop. If breadth is what you need, choose Redgate.

Enterprise maturity

Redgate has been at this since 1999, with a large proven install base, established procurement, support, and compliance posture. MyDBA is newer — capable, but not decade-hardened.

SQL Server depth

Redgate's SQL Server tooling and ecosystem are best-in-class. If SQL Server is your centre of gravity and Postgres is the side workload, Redgate's heritage is a real advantage.

One vendor, whole stack

Redgate spans monitoring, source control, deployment and more. If you want a single vendor across the database lifecycle, that breadth is something a focused tool like MyDBA doesn't try to match.

What it costs

Free for one primary. Pro priced per instance.

MyDBA Pro costs £50 per primary instance and £25 per replica instance each month. Every monitored instance counts, including your first primary and development or staging systems.

All databases and team access included

Free includes one primary or standalone instance with no replicas, all its databases, all non-AI features and team access. Both Free and Pro include shared managed and self-hosted collectors; dedicated managed collectors cost £50 per machine per month and do not increase monitoring allowances. Both plans keep 30 days of history. Free sampling slows after the first seven days.

Worked exampleA primary + 2 replicas: MyDBA Pro costs £100/month before dedicated collector charges. For another provider, compare your actual quote, replica licensing, contract duration, support and currency alongside the features you need.

AI is optional: Free and Pro support OpenRouter, OpenAI or Anthropic customer keys with no MyDBA monthly quota (provider charges apply). New Pro includes 20 platform-funded operations per organization per UTC calendar month.

See the current Free and Pro plans and calculate your cost →

The bottom line

Pick Redgate for the estate. Pick MyDBA when Postgres is the thing that matters.

It's a breadth-vs-depth choice. Redgate wins when you need many engines in one place. But because MyDBA does nothing but Postgres, it goes far deeper — advice tailored to PG internals, incident post-mortems, anomaly detection, verified tuning and extension monitoring that an estate-wide view isn't built to do. Start free on one primary instance — no card.