Turn your servers into a complete cloud.
From raw physical servers to a highly available, multi-tenant cloud. The Tres orchestrator and operator panel run virtualization, networking, storage and managed services for you — instead of spending months stitching together open-source components, you rent a ready-to-run, supported cloud orchestrator.
Everything a cloud provider needs — ready to run
From virtualization to high availability, from networking to managed services, from billing to white-label. You don’t integrate separate tools; everything comes under a single orchestrator and a single panel.
Virtualization & High Availability
Networking & Security
Managed Services & Catalog
Operations & Commercialization
The cloud stays up even if a single node fails
High availability isn’t a bolt-on module; it’s at the core of the orchestrator. The cluster balances and heals itself.
HA Cluster
Nodes run as a single cluster and resources are pooled. Single points of failure disappear; capacity grows with every node added to the cluster.
DRS · Dynamic balancing
Load is distributed across nodes automatically. As hot spots form, virtual servers are moved to less-loaded nodes via live migration — without downtime.
Failover
If a node goes down, the virtual servers on it are automatically restarted on healthy nodes. Recovery needs no manual intervention.
Building a cloud from scratch takes months. We deliver it ready to run.
Instead of stitching together open-source hypervisor, scheduler, networking, storage and billing components yourself and taking on their maintenance, you get a single platform that’s integrated, kept up to date and supported 24/7.
| Capability | If you build it yourself | Tres Orchestrator |
|---|---|---|
| Virtualization & orchestration | ✕You integrate separate hypervisor, scheduler, networking and storage | ✓A single integrated engine, ready to run |
| High availability (HA/DRS/Failover) | ✕Manual setup, constant tuning and testing | ✓Enabled out of the box |
| Networking, VPC and firewall | ✕Separate components, manual rule management | ✓One click from the panel, global policy |
| DBaaS & Kubernetes | ✕You install, patch and manage it yourself | ✓Included as a managed service |
| Panel · API · Terraform · AI | ✕None, or you write it yourself | ✓All ready and consistent |
| Metering, billing & white-label | ✕Requires a separate system | ✓Metering, billing and your own brand included |
| Updates, patching & support | ✕Entirely your responsibility | ✓By Tres, supported 24/7 |
| Time to go live | ✕Weeks / months | ✓Days, with a guided rollout |
On our cloud, on dedicated hardware, or on your own devices
The orchestrator is the same — the panel, API and AI stay exactly the same. The only thing that changes is where the hardware is and who owns it.
Public Cloud
Instant access, no hardware required.
Hosted Private Cloud
Dedicated hardware in a Tres data center.
On Your Own Hardware
We install the orchestrator on your own devices.
For details on hosted and on-your-hardware deployments, see the Private Cloud page.
Manage it just by talking, if you like
Alongside the panel, REST API and Terraform, a live MCP server is included too. You connect assistants like Claude to your infrastructure and perform operations such as creating or scaling servers in natural language — every command is written to the audit log.
API · Terraform · MCP →Snapshot taken (snap-4f2a). web-01 resized to TRC-8. ✓ Written to the audit log.
The hard part of building a cloud isn’t powering on servers, it’s operating them as a single system: virtualization, high availability, networking, storage, backup, managed services, metering and billing. Assembling these from open-source parts yourself means months of integration and a constant maintenance burden. Tres Platform removes that burden: you rent a ready-to-run, integrated and supported cloud orchestrator.
What Is a Cloud Orchestrator?
A cloud orchestrator is the control layer that manages a pool of physical servers as a single cloud. It coordinates compute (virtual servers), storage, networking and managed services with one logic; creating servers, taking snapshots, scaling, configuring networks, backup and billing all run through the same orchestrator. The Tres orchestrator runs on an engine we developed ourselves — because we don’t depend on a third-party hypervisor license, no license cost is passed on to your price and new features are added quickly.
When you rent it together with the operator panel, your raw servers turn into a multi-tenant, highly available cloud within a few days. Your end users open virtual servers, set up databases and create Kubernetes clusters from their own panels; you manage the nodes, clusters, network policies and billing from the operator panel.
High Availability: HA Cluster, DRS and Failover
What separates an enterprise cloud from “just virtual machines” is high availability. In the Tres orchestrator this isn’t a bolt-on module but part of the core:
- HA Cluster: Nodes run as a single cluster and resources are pooled. Single points of failure disappear; capacity grows with every node added.
- DRS (Dynamic Resource Scheduling): Load is balanced across nodes automatically. As hot spots form, virtual servers are moved to less-loaded nodes via live migration — without service interruption.
- Failover: If a node goes down, the virtual servers on it are automatically restarted on healthy nodes; recovery needs no manual intervention.
Tip: Setting up DRS and Failover yourself requires constant tuning and testing on a running cluster. In the orchestrator both come switched on out of the box; they activate automatically when you add a node to the cluster.
Networking, VPC and Security
In a multi-tenant cloud, network isolation and security are the most critical and most laborious part of the job. The Tres orchestrator makes this manageable from the panel: VPC / private networks for each tenant, Floating IP and IP pool management (including BYOIP and BGP), a per-server distributed firewall and security groups. There’s also a global network policy applied to all virtual servers: rules such as ARP flood and IP spoofing protection, DHCP and metadata access control, and outbound SMTP blocking are enabled from a single place. A managed load balancer is provided from the same panel too.
Not Just Virtual Machines: Managed Services
Most “cloud on your own hardware” solutions offer only virtualization; they don’t cover managed services like databases and Kubernetes. The Tres orchestrator manages these services too:
- Managed Database (DBaaS): MySQL, PostgreSQL, MariaDB, MongoDB and Valkey install, back up and scale with one click.
- Managed Kubernetes: The control plane and worker pool are provided by the orchestrator; the cluster can run in the public cloud or on your own hardware.
- Application catalog: Ready-made apps and databases install with one click; operating system images and an ISO library come in the catalog.
So whichever delivery model you choose — public cloud, hosted private cloud or your own hardware — you get access to a full service catalog from a single platform.
Operations, White-label and Billing
Building a cloud isn’t enough — you have to operate it and, if needed, sell it. The Tres orchestrator is complete on the operator side too: per-resource metering and billing, an audit log, roles and permissions, and location/region management. With white-label support you offer the cloud under your own brand and domain — each brand gets its own login and runs separately on the same orchestrator. This suits both service providers that want to offer their own cloud/hosting service and large organizations that provide an internal cloud.
Tip: Thanks to white-label and multi-tenancy, you can also use the platform to “become your own cloud provider”: you open sub-tenants, define quotas and billing for them, and serve them under your own brand.
Automation: API, Terraform and AI/MCP
Modern infrastructure is managed with code. The entire orchestrator can be managed programmatically with the REST API, defined as infrastructure-as-code (IaC) with Terraform and integrated into CI/CD pipelines. On top of that, a live MCP server is included: you connect AI assistants like Claude to your infrastructure and have operations carried out in natural language. The panel, API, Terraform and AI all run on the same orchestrator; whichever you use, the result and the audit trail are consistent.
Three Delivery Models
The same orchestrator is delivered in three ways; the panel, API and AI are exactly the same in all three.
| Model | Hardware | Best suited for |
|---|---|---|
| Public Cloud | Tres cloud pool | Fast start, variable load, pay-as-you-go |
| Hosted Private Cloud | Dedicated to you in a Tres data center | Compliance, isolation, predictable performance |
| On Your Own Hardware | Your own devices | Data must stay with you, existing hardware investment |
As you start on the public cloud and grow, when you move to a hosted private cloud or your own hardware, your panel, API and automation scripts stay the same; you don’t learn a new system. For details, see our Private Cloud page.
Who Is It For?
It’s a fit for organizations that want to build a highly available internal cloud on their own hardware; scenarios where data must physically stay within the organization; service providers that want to offer a cloud/hosting service under their own brand; and teams that want to manage their infrastructure from a single panel, with code and AI.
Frequently asked questions about the cloud orchestrator
Tres Platform is an orchestrator and operator panel that runs your pool of physical servers as a single cloud. You manage virtualization, high availability (HA/DRS/failover), networking, storage, backup, DBaaS, Kubernetes, metering and billing from a single platform. It comes ready to run and supported; you don’t assemble the components yourself.
Yes. The orchestrator is installed on your own hardware; your raw servers turn into a multi-tenant, highly available cloud and the data stays entirely with you. Alternatively, it can run on dedicated hardware in a Tres data center or on our public cloud.
HA Cluster runs the nodes as a single cluster and pools their resources. DRS balances load across nodes automatically and cools down hot spots with live migration. Failover automatically restarts the virtual servers from a node that goes down on healthy nodes. All three come at the core of the orchestrator, out of the box.
You could, but integrating the hypervisor, scheduler, networking, storage, backup and billing is a project that takes months and a constant maintenance burden; keeping HA/DRS/failover working takes separate expertise. The Tres orchestrator delivers all of this as a single platform that’s integrated, kept up to date and supported 24/7.
DBaaS (MySQL, PostgreSQL, MariaDB, MongoDB, Valkey), managed Kubernetes, block storage, snapshots and backup, a load balancer and a one-click application catalog. These services work the same way not only in the public cloud but also in a deployment on your own hardware.
There are VPC / private networks for each tenant, Floating IP and IP pool management (BYOIP/BGP), and a per-server distributed firewall and security groups. In addition, a global network policy applied to all virtual servers lets you enable ARP flood, IP spoofing, DHCP, SMTP and metadata protection from a single place.
Yes. The orchestrator is multi-tenant and white-label: you offer the cloud under your own brand and domain, and each brand gets its own login. Because metering and billing are included, you can offer your own cloud/hosting service.
Yes. All resources are managed with the REST API, defined as infrastructure-as-code with Terraform and integrated into CI/CD. With the live MCP server you can also connect assistants like Claude and manage it in natural language. The panel, API, Terraform and AI produce the same result, and every operation is written to the audit log.
No. Because the three delivery models use the same orchestrator, your panel, API and automation scripts stay the same during the transition. You don’t learn a new system; this also reduces the risk of provider lock-in.
The on-your-hardware and hosted models are delivered as a guided rollout; the architecture is designed together and typically goes live within days. In the first meeting we produce a draft architecture and a cost range.
Tres infrastructure is operated in Tier III+ data centers; it offers ISO 27001, PCI DSS, KVKK and 5651 compliance along with a 99.9% SLA commitment. In the on-your-hardware model, data physically stays in your own facility; every operation is written to the audit log.
Let’s build your cloud together
Tell us about your hardware and your needs; in the first meeting we’ll produce a draft architecture and a cost range. Or try the public cloud right away.