Gerrit is built for code review.
Active MultiSite makes it work across global teams.
Where Gerrit Active MultiSite Excels
Code reviews are slowing down across regions
Developers work against a nearby Gerrit server while review data stays in sync globally.
You’re adding new development sites
Bring new locations online without rebuilding your Gerrit infrastructure.
One server outage stops development
Multiple writable sites keep teams working even if a location goes offline.
Keep Gerrit Fast, Reliable,
And Easy to Scale.
Stay Online in Outages
- High Availability: Multiple Gerrit servers work together so there’s no single point of failure.
- Built-in Disaster Recovery: Every site maintains a live copy of repositories and review data.
- Automatic Catch-Up: Sites that go offline automatically sync the changes they missed when they reconnect.
Work at Local Speed
- Fast Local Clones: Developers interact with the nearest Gerrit server for faster repository access.
- Globally Consistent Reviews: Comments, approvals and permissions stay aligned across every site.
- Reliable Pushes: Configurable replication policies balance speed, redundancy and recoverability.
Grow Your Infrastructure
- Single Server Consistency: PAXOS technology guarantees ensures Gerrit servers operate together as one environment.
- Selective Replication: Choose which repositories are available at each location.
- Add New Sites Easily: Bring new development locations online without rebuilding your infrastructure.
What ActiveMultiSite Makes Possible.
Active MultiSite works behind the scenes to keep every Git, Gerrit, and Subversion server in sync while developers work locally.
Here are the core capabilities that make that possible.
Guaranteed consistency
PAXOS technology ensures all Gerrit sites stay synchronized after every change, so teams work from the same repository and review state across locations.
Automatic catch-up after outages
If a site goes offline, Gerrit AMS automatically synchronizes the changes it missed when it reconnects — no manual recovery required.
High availability
Multiple writable sites operate together, removing the single point of failure common in centralized Gerrit deployments.
Built-in disaster recovery
Every site maintains a live copy of repositories and review data, allowing development to continue if a location goes down.
Selective replication
Choose which repositories are replicated to each site to support access policies and reduce infrastructure overhead where full replication isn’t required.
Fast local clones
Repository reads are served from the nearest site, giving developers fast access to code, even in globally distributed teams.
Reliable pushes
Configurable redundancy policies balance performance and resilience across sites.
Globally consistent reviews
Comments, approvals and review metadata stay synchronized, so teams see the same review state everywhere.
How Global Engineering Teams Scale
Customer Success Stories
Scalability solution for a major phone manufacturer
Challenge
- Horizontal scalability issues with 9,000+ users and 11,000+ repositories.
- The Gerrit "high availability" plugin, which supports multiple read-only sites, was insufficient to handle the high access load on the read-write server.
Results
- Implemented a DConE solution, deploying 7 multi-primary sites across various locations to distribute and manage the end-user load effectively.
- Continued use of the Gerrit "high availability" plugin to offload simple Git clones and fetches.
Efficient global collaboration for an electronics manufacturer
Challenge
- Need for efficient collaboration across global development sites.
- Requirement for fast access to large files during Git checkouts.
Results
- Enabled efficient global collaboration by replicating Git LFS storage across development sites. This ensures large files are latency-close, allowing for quick downloads and improved productivity.
- Allowed for cost-effective management by locating Git project repositories only where needed, optimizing server usage for specific use cases.
Robust, no-failure software development for an automotive manufacturer
Challenge
- Eliminate single points of failure in a single data center setup.
- Ensure seamless maintenance without downtime.
Results
- Deployed 3 primary-primary sites with single-copy identical data. This setup allows any one site to be taken down for maintenance while the other two remain fully operational in read-write mode.
- Prevents server stalls by ensuring data is not locked in a down site, maintaining smooth operations.
- Proxy server positioned in front of the three sites to ensure seamless availability and access.
Resources
Datasheet
Published July 2025
User Guide
v3.9 (latest)
Git Binaries
Git Version | Operating System (OS) | Download |
|---|---|---|
2.39.1 | CentOS 6 - Git 2.39.1 | Download |
2.39.1 | CentOS 7 - Git 2.39.1 | Download |
2.39.1 | RedHat Enterprise Linux 6 - Git 2.39.1 | Download |
2.39.1 | RedHat Enterprise Linux 7 - Git 2.39.1 | Download |
2.39.1 | SUSE Linux Enterprise 11 - Git 2.39.1 | Download |
2.39.1 | SUSE Linux Enterprise 12 - Git 2.39.1 | Download |
2.39.1 | Ubuntu 12.04 - Git 2.39.1 | Download |
2.39.1 | Ubuntu 14.04 - Git 2.39.1 | Download |
2.31.1 | Windows x64 | Download |
2.31.1 | Ubuntu 20.04 Focal | Download |
Keep code reviews fast — wherever your teams work
As you grow across regions and development sites, code review infrastructure needs to keep pace. See how Gerrit Active MultiSite helps scale your review environments while maintaining speed, resilience, and consistency.