Platform · Deployment architecture
Start on one server and grow without rebuilding.
Start with one 12Port instance for your evaluation. Run Secure Vault, Access Broker and Account Management on your own server, VM or cloud account, then expand to a highly available or geographically distributed deployment as your needs grow.
Six deployment models
The same install, from one server to many regions.
Grow with the same product, from your first brokered session to multiple sites and regions. Plan the database, network, capacity and recovery requirements for each stage.
Single node
Trials, labs, small teams
One server runs the application services with an embedded database for trial, test or smaller deployments. Access the web console and supported web sessions over HTTPS, or use native RDP and SSH clients through the configured gateways.
Extended
Production, internet-facing
Add a supported external database and, where needed, a load balancer. HTTPS can use port 443; native RDP and SSH require their own configured services and ports.
High availability
Multiple application nodes
Use two or more application nodes with a shared external database and a suitable load balancer. Configure sticky sessions for web sessions and plan database and load-balancer availability as part of the design.
Disaster recovery
Off-site backup that can take over
Use a recovery node with a separate database and a defined replication or backup-and-restore process. Configure traffic switching and test recovery. Node-to-node replication can synchronize data collected during an outage back to the main node.
Multi-region
Users and assets in more than one place
Deploy nodes with independent databases in different datacenters or clouds. Replicate configuration and operational data continuously or on a schedule. Cross-region access and failover depend on network reachability and the configured routing.
Multi-site with peer nodes
Isolated networks, plants, client sites
A main node stores configuration, keys, assets and audit data. Peer nodes provide access to endpoints and execute configured jobs at remote sites. Connect them directly or through configured peer tunnels.
“We moved over 150 users off of our previous PAM platform in just a few weeks and took advantage of 12Port’s great support and implemented a highly available multi-region PAM.”
12Port customer, global consumer products company
Replication
Replication built into the application, not bolted onto the database.
Each 12Port node keeps a journal of every create, update and delete across configuration, management objects and logs. The replication service ships that journal to its peers over the REST API on HTTPS. Because it replicates objects rather than database pages, the two sides can run different databases, and a peer in another cloud needs nothing more than an HTTPS path.
- Per tenant: replicate the tenants that need it, leave the rest local.
- Push or pull, chosen by which side can open the connection, so it works through one-way firewalls.
- Active data (configuration, assets, sessions, requests) moves immediately or in the background; logs and reports move on a schedule.
- Active-active or active-passive between nodes, with real-time, deferred or synchronization modes for multi-region deployments.
- Star topologies: a main node can sync only what it originated with each peer.
- Replication uses API tokens with the required permissions. Choose which tenants and data to replicate as part of your architecture.
What it gives you
Flexible topology. Choose shared-database high availability or independent-database replication according to your availability and recovery requirements.
Less to manage. One product, one configuration, one audit trail, replicated. Adding a region is a node and a replication profile.
Plan and test recovery. Failover requires reachable target systems, suitable routing and an available data copy. Recovery time and data loss depend on the architecture and replication or backup schedule.
Reach
Reach remote networks with planned connectivity.
Add a peer node where the central 12Port server cannot directly reach endpoints, such as another location or an isolated OT network. Where inbound connections are blocked, a configured outbound SSH tunnel can provide connectivity.
Peer nodes
A peer node brokers supported connections and runs configured discovery and account-management jobs near remote endpoints. It needs a network path to the main node, directly or through a peer tunnel. A fully air-gapped environment requires a deployment within that environment.
Peer tunnels
When inbound connections to the peer site are blocked, the peer can establish an outbound SSH tunnel to an accessible SSH server on the main host, normally on port 22. Firewall policy must permit that connection. One peer can maintain tunnels for multiple tenants or main nodes.
Relay nodes
A relay node, a caching peer node, relays sessions near users and endpoints to optimize geographically distributed traffic, while replication shares configuration and audit data with the main node. Plan where replicated data and backups are stored.
Runs where you need it
Self-hosted on supported Windows and Linux servers.
Host 12Port on your own VM infrastructure or in your own AWS, Azure, Google Cloud or other cloud account. Choose where your PAM data is stored to support your data residency and sovereignty requirements. Include databases, replicas, backups and external integrations in that plan.
Next step
Plan it with us, or start with one server.
Start in your own environment and grow the deployment as your needs evolve. Talk to an engineer about your sites, regions and recovery targets.