---
title: "Multigres private alpha: automatic failover for Supabase Postgres"
description: "Multigres, an open-source operating system for Postgres built by the Vitess team, has entered private alpha on the Supabase platform. It provides automatic failover across three nodes in different availability zones, guaranteeing no acknowledged writes are lost when the primary fails. Failover is treated as a consensus problem using Postgres synchronous replication, so the system acknowledges a write only after it has been replicated to enough nodes to survive a failure. Built-in connection pooling eliminates the need for a separate pooler and deduplicates prepared statements across clients. The alpha is free for up to two projects per organization on paid plans, but it is not covered by the normal SLA, does not support Realtime or Point-in-Time Recovery, and cannot be enabled on existing projects."
---

# Multigres private alpha: automatic failover for Supabase Postgres

[Supabase](https://yomu.fyi/company/supabase) · aditya\_maruvada,manan\_gupta · Oct 29, 2026

**Type:** Announcement

## Summary

Multigres, an open-source operating system for Postgres built by the Vitess team, has entered private alpha on the Supabase platform. It provides automatic failover across three nodes in different availability zones, guaranteeing no acknowledged writes are lost when the primary fails. Failover is treated as a consensus problem using Postgres synchronous replication, so the system acknowledges a write only after it has been replicated to enough nodes to survive a failure. Built-in connection pooling eliminates the need for a separate pooler and deduplicates prepared statements across clients. The alpha is free for up to two projects per organization on paid plans, but it is not covered by the normal SLA, does not support Realtime or Point-in-Time Recovery, and cannot be enabled on existing projects.

## Context

The post explains that current Supabase projects write to a single Postgres primary, which becomes unavailable during hardware faults or zone outages. Teams needing high availability build their own solutions with replicas, failover managers, and poolers. Multigres is an open-source operating system for Postgres from the Vitess team, now available as a private alpha on Supabase, managing the cluster with automatic failover, connection pooling, replication, and backups.

## Approach / What changed

Multigres treats failover as a consensus problem using Postgres synchronous replication. It only acknowledges writes after they are replicated to enough nodes to survive a failure, ensuring no committed writes are lost when a replica is promoted. The Multigres gateway can hold incoming requests during failover to avoid connection errors. Built-in connection pooling parses each request and tracks session state to handle every query without a pooling mode selection; it also deduplicates prepared statements across clients and provides fair-share connection budgets per database user.

## Takeaways

- Multigres uses a consensus protocol based on Postgres synchronous replication to guarantee that acknowledged writes survive failover, unlike traditional failover tools that treat it as a monitoring problem.
- Built-in connection pooling automatically handles query parsing and session state, eliminating the need to choose a pooling mode and deduplicating prepared statements across clients.
- During the private alpha, Multigres is free for up to two projects per organization on paid plans, but it is not covered by the SLA, lacks Realtime and Point-in-Time Recovery, and cannot be enabled on existing projects.

## FAQ

### What is Multigres?

Multigres is an open-source operating system for Postgres built by the Vitess team. It manages a Postgres cluster around your standard Postgres data, handling connection routing, pooling, replication, failover, and backups.

### How does Multigres ensure no data loss during failover?

Multigres uses a consensus protocol on top of Postgres synchronous replication. Your application receives an acknowledgement for a write only after it has been replicated to enough nodes to survive a failure, so every acknowledged write exists on at least one surviving node and is not lost when a replica is promoted.

### Can I enable Multigres on an existing Supabase project?

No, Multigres must be chosen at project creation and cannot be enabled on an existing project. There is currently no managed path back to a standard Postgres project.

**Tags:** [High Availability](https://yomu.fyi/topic/high-availability), [Multigres](https://yomu.fyi/topic/multigres), [Postgres](https://yomu.fyi/topic/postgres), [Supabase](https://yomu.fyi/topic/supabase)

- Source: [Supabase](https://supabase.com/blog/multigres-private-alpha)
- Source URL: https://supabase.com/blog/multigres-private-alpha
- Ingested by Yomu: 2026-10-29T18:01:10.664Z

[Read original post](https://supabase.com/blog/multigres-private-alpha)
