High availability clustering in ProVide #
ProVide Cluster lets you run two or more ProVide Servers together as a single service. Users see one system. Behind the scenes, traffic is shared across multiple nodes, and configuration, users, permissions, links, and certificates are kept automatically in sync.
The result: higher capacity, built-in redundancy, and the ability to take a server out of service without taking the service down.
Added together with one or two ProVide Gateways, the cluster also gets intelligent load balancing, health monitoring, and a secure single-entry point for external traffic. With two Gateways, ProVide can perform rolling updates of both Gateways and nodes without interrupting the service.
The three levels: Cluster, Cluster + Gateway, Cluster + 2 Gateways #
Clustering was introduced in ProVide release 2026-04 (Iron).
ProVide’s clustering scales with your needs. You can start small and add gateways when the use case calls for it.
Cluster (nodes only) Two or more ProVide nodes work together. Configuration is synchronized automatically: users, groups, permissions, sharing links, certificates, and settings. If one node goes down, the others continue serving traffic. Updates are scheduled manually, one node at a time.
Cluster + 1 Gateway A ProVide Gateway is added in front of the cluster. The Gateway handles all external traffic (HTTPS, SFTP) and distributes connections across nodes based on real-time load, health checks, and sticky sessions. Only the Gateway is exposed to the internet, so the ProVide nodes can run on a protected internal network. The Gateway itself is a single point of failure at this level.
Cluster + 2 Gateways A second Gateway eliminates the single point of failure in the front and unlocks fully automated rolling updates. ProVide does a full update automation and the service stays available throughout, no maintenance window required.
When do you need it? #
ProVide Cluster is built for organizations where the file transfer service is business critical. Common scenarios:
24/7 operations in healthcare, finance, logistics, e-commerce, and public sector, where planned downtime is not acceptable.
High traffic volumes with many concurrent SFTP or HTTPS transfers, where scaling horizontally across nodes is preferred over buying a single larger server.
SLA commitments to customers or partners, where every minute of downtime carries a financial or reputational cost.
Compliance-driven environments where business continuity is a requirement in contracts or regulations.
Geographic redundancy with nodes in two data centres or regions that can take over for each other.
Security policies that require external traffic to enter through a DMZ component rather than reach the file server directly.
If none of those apply, a single ProVide Server is the right choice. Clustering is an add-on for organizations that need the extra layer.
What’s required to get started #
You can start at any level and grow into the next when the need is there.
Cluster (nodes only): at least 2 ProVide nodes.
Cluster + 1 Gateway: 2 ProVide nodes + 1 Gateway.
Cluster + 2 Gateways: 2 ProVide nodes + 2 Gateways.
Technical requirements in short:
ProVide nodes run on Windows Server. Physical, virtual, or cloud based.
Gateways run on Windows or Linux. The Gateways do not need to use the same operating system as the nodes.
Server license: Maxi or Mega
Cluster licensing: one Cluster license per node, or a Cluster Unlimited license.
Gateway licensing: one per Gateway, or a Gateway Unlimited license.
Network: nodes need to reach each other on the cluster port (default 3145, TLS recommended). Each node connects to the Gateway which means that only the Gateways need to be exposed to the internet.
For the full installation walkthrough, see [How do I set up an SFTP cluster with load balancing and failover?]
For details on how zero-downtime updates work, see [How do rolling updates work for zero-downtime SFTP server maintenance?]
Note: Technically the Gateway feature works with a single gateway, but in that case the gateway itself becomes a single point of failure. For customers who need true high availability, two Gateways is the answer.
