UPVALE / GUIDES
Getting started
Connect your servers, create a project, and run your first app, database, or scheduled job.
On this page
Upvale runs container applications, databases, and scheduled jobs on the servers you connect. You manage them from the dashboard. Upvale hosts and maintains the control plane; you do not install a separate Upvale server.
Your first deployment
- Open Fleet → Add cluster. Name the cluster and choose how its nodes communicate. Use Private VPC / VPN for an existing private network, Public / mixed for directly reachable addresses, or Managed WireGuard for a private mesh managed by Upvale.
- Enroll a server with the command shown by Add node. Run it on that server with the required root privileges. Wait until the node is online and has its Swarm manager or worker role. Cluster setup covers the full process and network requirements.
- Open Workloads → New project, enter a name and assign a cluster. The project starts with a Production environment. You can also assign its cluster later in the project's settings.
- Open the project and use Add Service. Choose Application, Database, Scheduled job, or Compose stack.
- Follow the matching guide below. For applications and databases, saving configuration and deploying it are separate steps. Follow the queued operation until it succeeds.
| I want to… | Start here |
|---|---|
| Run a web app or worker from a container image | Applications: simple setup |
| Run PostgreSQL, Redis, or MongoDB | Databases: simple setup |
| Run PostgreSQL with failover and optional read replicas | Managed PostgreSQL HA — minimum three active nodes |
| Run a command on a schedule | Scheduled jobs |
| Bring an existing Compose file | Compose |
| Connect a developer's laptop to a private database | Developer VPN |
For a first web deployment, a public nginx image on container port 80 is enough. Add a hostname that you control, point its DNS at your ingress address, and deploy. You can add resource limits, secrets, volumes, and custom traffic rules when you need them.
How things fit together
| Name | Meaning |
|---|---|
| Organization | Your team's projects, servers, shared credentials, and permissions. Switch organizations from the account menu. |
| Cluster | A group of servers with shared workload networking and ingress. |
| Node | One server enrolled in a cluster. A manager coordinates Swarm; workers run assigned tasks. Managers can also run workloads. |
| Project | Related applications, databases, and jobs. |
| Environment | A resource scope inside a project, assigned to one cluster. The current UI creates Production; use separate projects when you need separate staging and production setups. |
| Resource / service | An application, database, or job and its saved configuration. |
| Revision | A saved version of a resource's desired configuration. Saving does not prove that version is running. |
| Operation | Background work, such as deployment, node updates, backup, or restore. Open it to see progress and failed steps. |
| Replica / task | A running instance of a service. Application replicas and PostgreSQL standby servers serve different purposes. |
A project or environment groups resources; it is not automatically a network isolation boundary. Choose networking deliberately when services should not be able to reach one another. See Networking.
Before you add advanced settings
Start with a working image, a cluster, and the smallest configuration your app needs. Check its logs after deployment. Then add:
- Registry credentials for private images.
- Secrets for passwords and connection strings.
- Persistent storage and backup destinations for data that must survive replacement containers.
- Routes and traffic rules for domains, HTTPS, path changes, authentication, and limits.
- Alerts for failures you want to hear about.
- Team roles so people have the access they need.
Find a screen or fix a problem
The dashboard and screen guide maps every customer-facing area to its manual. Each workload guide starts with a short setup, then describes its tabs, advanced settings, and common problems.
Use Troubleshooting when a node, deployment, route, backup, or VPN connection is not working. Controls can be hidden or disabled by your role or subscription; check Team and account if something is missing.