EDB Postgres Distributed (PGD)
For the workloads that cannot go down — EDB Postgres Distributed (PGD) runs active-active, geo-distributed Postgres clusters at up to 99.999% availability, with no failover gap and even online major-version upgrades, on open Postgres.
Buy through TechBag
Same software. Better outcome — at a lower cost.
How it’s rated
Full scoreboard ↓Quick answer
The EDB Postgres platform family
This page covers Distributed HA (PGD) — extreme availability. The rest of the portfolio:
Quick facts
30-second orientation- Product
- EDB Postgres Distributed (PGD)
- Vendor
- EDB (EnterpriseDB) — independent, Bain Capital-backed
- What
- Active-active, geo-distributed Postgres HA
- Availability
- Up to 99.999% (‘five nines’)
- The model
- Multiple nodes all accept writes (not single-primary)
- Also
- Low latency for geo-distributed apps
- Key feature
- Online maintenance & major-version upgrades
- Eliminates
- Failover gaps AND planned-downtime windows
- For
- Mission-critical, latency-sensitive workloads
- In India via
- TechBag — quotes, PoCs, GST invoicing, Tier-1 support
Understand active-active HA before you buy it
Most product pages skip this. We start here — so you buy a capability, not a buzzword.
What is PGD?
EDB's active-active, geo-distributed Postgres high availability — multiple nodes all accept writes, delivering up to 99.999% uptime.
Plus online maintenance and even major-version upgrades without downtime.
Single-primary HA vs active-active PGD — the honest table
What consolidation actually replaces, dimension by dimension.
| Dimension | Single-primary (failover gap) | PGD (active-active, five nines) |
|---|---|---|
| HA model | Single primary + standbys | Active-active, all nodes write |
| Failover | Gap (seconds–minutes) | No gap — others keep serving |
| Availability | Lower nines | Up to 99.999% |
| Writes | Only the primary | All nodes |
| Latency (geo) | Round-trip to primary | Local writes, low latency |
| Maintenance | Downtime windows | Online, no downtime |
| Major upgrades | Disruptive downtime | Online, no downtime |
| vs Oracle RAC | Proprietary, expensive | Open Postgres, no lock-in |
Extreme HA on open Postgres — active-active is complex, for workloads that genuinely need it; standard needs suit single-primary.
The five pieces of the platform
Vendors love diagrams; buyers need to know what they’re actually operating. Here’s the whole platform, demystified.
Active-Active Cluster
Multiple database nodes all accept writes simultaneously — no single primary to fail. If a node goes down, the others keep serving, with no failover gap.
Geo-Distribution
Nodes can be spread across geographic locations — users write to their nearest node, giving low latency for geographically-distributed applications.
Conflict Resolution
Manages the write conflicts inherent in multi-node writes — the sophisticated machinery that makes active-active Postgres work reliably.
Online Operations
Online maintenance and even online major-version upgrades — patch, maintain and upgrade without taking the database down, eliminating planned-downtime windows.
Fast Failover
Automatic, fast handling of node failures — the cluster keeps serving, delivering up to 99.999% availability.
One active-active Postgres cluster across locations — conflicts resolved, nodes failed over, upgrades done online.
Twelve capabilities. Availability, distributed, operate.
PGD runs active-active Postgres — all nodes write, no single primary to fail — delivering five-nines availability and zero-downtime upgrades.
Active-Active Writes
Multiple nodes all accept writes — no single primary to fail, no write bottleneck.
Up to 99.999% Uptime
‘Five nines’ availability — about five minutes of downtime per year, for workloads that can't go down.
No Failover Gap
If a node fails, others keep serving — no primary-promotion gap where the database is unavailable.
Geo-Distributed
Nodes across locations — users write to their nearest node, low latency for distributed apps.
Low Latency
Local writes for geographically-distributed users — not a round-trip to a distant primary.
Conflict Resolution
Handles multi-master write conflicts reliably — the machinery that makes active-active work.
Online Maintenance
Patch and maintain the database without downtime — no more planned-maintenance windows.
Online Major Upgrades
Even major-version upgrades without taking the database down — a genuinely rare, valuable capability.
Read/Write Scale
Distribute load across nodes — availability and scalability together.
On Open Postgres
Extreme HA on true open Postgres — runs anywhere, no lock-in.
EDB Enterprise Support
Enterprise support for your most critical HA workloads — the safety net that matters most here.
On Extended/Advanced Server
Built on EDB's enterprise Postgres — the HA layer for your migrated or Postgres-native workloads.
Watch EDB Postgres Distributed in action
Active-active, geo-distributed Postgres HA and five-nines availability.
Introducing EDB Postgres AI
EDB's sovereign data-and-AI platform built on Postgres, introduced by EDB.
EDB Postgres AI: Enterprise-Grade Postgres
What makes EDB's Postgres enterprise-grade — HA, security, support.
Postgres Architect Answers AI & Data Questions
An EDB Postgres architect fields real AI and data questions.
Want a live, India-context walkthrough for your environment?
Book a guided demo →A single primary can fail. Active-active doesn’t.
Here’s what genuinely sets EDB Postgres Distributed (PGD) apart from the alternatives.
Active-active beats single-primary HA
Standard database HA has one primary and passive standbys: failover leaves a gap of seconds to minutes, and standbys don’t serve writes. PGD is active-active — multiple nodes accept writes, so there’s no single primary to fail and no gap. That’s why PGD reaches up to 99.999%, five nines.
Five nines: for workloads that can't go down
99.999% availability means about five minutes of unplanned downtime per year — what core banking, telecom infrastructure, critical government services and payment processing need, where downtime is a crisis with financial, regulatory and reputational cost. PGD delivers it on open Postgres.
Low latency for geo-distributed apps
PGD nodes can be spread across locations and all accept writes, so users write to their nearest node instead of a distant primary. A global application gets low latency and resilience against any single location failing — which single-primary architectures can’t provide.
Upgrade without downtime — genuinely
PGD supports online maintenance and even online major-version upgrades, so you can patch and upgrade without taking the database down. Planned-maintenance windows and disruptive major-version upgrades disappear — continuous availability, planned and unplanned.
Extreme HA, on open Postgres
PGD delivers active-active, five-nines availability on open PostgreSQL, running on-prem, on any cloud or hybrid, built on EDB’s enterprise Postgres (Extended or Advanced Server). You get mission-critical resilience without the cost and lock-in of proprietary HA like Oracle RAC.
The honest positioning
PGD’s five nines and online upgrades are real strengths, but active-active is complex — conflict resolution and app design matter — so it’s for workloads that need extreme HA. Alternatives: single-primary Postgres HA, Oracle RAC, or distributed SQL (Yugabyte, CockroachDB). TechBag scopes whether you need it.
The numbers behind the platform
What your EDB Postgres Distributed (PGD) journey looks like
Availability scoping
Your mission-critical workloads, availability SLA (do you truly need five nines?), and geo-distribution needs. TechBag scopes it free.
PGD PoC
Stand up an active-active PGD cluster; test failover (a node down, cluster keeps serving), geo-writes and an online upgrade.
HA design & app fit
Design the cluster topology; address conflict-resolution and application design for active-active; plan the migration to PGD.
Five-nines steady state
Mission-critical Postgres with active-active five-nines availability and zero-downtime upgrades. TechBag models it in INR/GST.
The review scoreboard
Modelled on Gartner Peer Insights structure. *Counts and breakdowns are illustrative pending verified review collection.
Quick poll — what’s driving your evaluation?
Talk to an advisor“Our core banking database can’t go down. PGD’s active-active model reached five-nines, with no single primary to fail.”
“Online major-version upgrades without downtime ended our planned-maintenance windows. True continuous availability.”
“Geo-distributed active-active lets global users write to their nearest node: low latency and resilience to any site failing.”
“Extreme HA on open Postgres, not proprietary Oracle RAC — mission-critical resilience without the licensing and lock-in. That combination sold it.”
“Active-active is complex — conflict resolution and app design need care — but for a workload that can’t go down, it was worth it.”
“Five minutes of downtime a year versus hours of failover gaps and maintenance windows with our old single-primary HA.”
“It builds on our EDB Postgres — the HA layer for our mission-critical workloads. Freedom from Oracle AND five-nines availability from one vendor.”
“For non-critical databases single-primary HA is simpler and fine — PGD is for the workloads where downtime is a crisis. We use it exactly there.”
Where everyone sits — the grids
Analyst firms bury this view behind paywalls, and review-site grids score satisfaction, not fit for your estate. So here’s TechBag’s synthesis of the active-active HA market — tap any vendor to see why it sits where it does.
TechBag Postgres HA Grid
Execution strength vs product vision — the classic market map, minus the paywall.
Active-active, five-nines Postgres HA — this page.
Availability × Openness
The grid nobody publishes — availability (active-active, five nines) vs openness (open Postgres, no proprietary lock-in).
Active-active + online upgrades + open — the corner it fills.
Positions are TechBag’s illustrative synthesis of public review-platform data and vendor documentation — not a reproduction of any analyst graphic. Verify before relying on it.
EDB Postgres Distributed vs the field
Single-primary HA, Oracle RAC and distributed SQL — honest lanes; the edge is active-active five-nines on open Postgres.
| Dimension | EDB Postgres Distributed | Single-primary Postgres HA | Oracle RAC | Yugabyte / Cockroach | No HA |
|---|---|---|---|---|---|
| HA model | Active-active, geo | Primary + standby | Active-active (Oracle) | Distributed SQL | None |
| Availability | Up to 99.999% | Lower | Very high | High | Low |
| Online upgrades | Yes (major too) | Downtime | Complex | Rolling | Downtime |
| Open / no lock-in | Open Postgres | Open Postgres | Oracle-locked | Postgres-compat | Depends |
| Best fit | Mission-critical Postgres that can't go down | Standard HA needs | Oracle RAC shops | Distributed-SQL adopters | Nobody critical |
Which HA approach fits you?
Honest fit signals — because the fastest way to lose your trust is to pretend one product wins every scenario.
Choose EDB Postgres Distributed (PGD) if…
- ✓Your workload genuinely cannot go down (five-nines needed)
- ✓You want active-active, geo-distributed Postgres HA
- ✓Online (incl. major-version) upgrades matter
- ✓You want extreme HA on open Postgres, not proprietary RAC
Single-primary Postgres HA if…
- ✓Standard availability needs — simpler, lower cost
Oracle RAC if…
- ✓You're deeply Oracle-committed (or migrate to PGD)
Yugabyte / Cockroach if…
- ✓You want a distributed-SQL database (different architecture)
No HA if…
- ✓Never for anything important — it's a single point of failure
EDB Postgres Distributed (PGD) is one of 28 databases & data tools products TechBag carries. The Databases & Data Tools guide narrows them to a shortlist and shows the reasoning. →
What does downtime cost you?
Drag the sliders (count mission-critical DB instances; IT-hour cost as a proxy for downtime cost per hour). Estimates model the annual downtime cost of single-primary HA (failover gaps + maintenance windows), with ~65% representative of the reduction from active-active five-nines with online upgrades — the avoided-downtime value for truly can’t-go-down workloads dwarfs the licensing. Illustrative.
Loaded cost = salary + overheads per productive hour. Illustrative only — your TechBag quote models your actual environment and modules.
Three ways to consume it
PGD prices as an add-on / subscription on EDB Postgres. TechBag models it vs Oracle RAC or downtime cost, in INR/GST.
PGD
Best for extreme HA
- Active-active, geo-distributed
- Up to 99.999% availability
- No failover gap
+ Online ops
Best for always-on
- Online maintenance
- Online major-version upgrades
- No planned-downtime windows
+ Open Postgres
Best vs Oracle RAC
- Runs anywhere, no lock-in
- On EDB enterprise Postgres
- TechBag scopes HA + migration
Buy it for less — TechBag pricing beats list
Whatever the list prices above, TechBag negotiates a significantly better deal — with GST-compliant INR invoicing and local support. Ask us for your discounted quote.
Get an India-ready quote
Tell us your requirements and current tools — we’ll model it against what you spend today.
The 8 questions to ask every Postgres HA vendor
Take this into your next vendor call — including ours.
Can your workload genuinely not go down? PGD is for mission-critical HA; single-primary suits standard needs.
Take a node down in a PoC — confirm the cluster keeps serving with no failover gap.
Test an online major-version upgrade — confirm it happens without downtime. A rare, valuable capability.
If your users are distributed, test local writes to the nearest node — low latency plus resilience.
Understand the active-active conflict-resolution and application-design implications — multi-master needs care.
Confirm it's on open Postgres — extreme HA without proprietary RAC lock-in.
Confirm EDB support — the safety net matters most for your most critical workloads.
Model PGD TCO vs Oracle RAC or downtime cost — TechBag quotes it in INR/GST.
Questions buyers ask
Ready to evaluate EDB Postgres Distributed (PGD)?
Scope a PGD PoC (test active-active failover and an online upgrade), or let a TechBag advisor design five-nines Postgres HA — in INR/GST.
Stats, ratings, review counts and pricing are illustrative and sourced from public materials; verify before purchase.