Deploy Soviez ERP on your own server.

Start from a supported Ubuntu server before piping the installer. You need root or sudo, a public hostname for Production, and enough CPU/RAM/disk for Docker, PostgreSQL, and ERP. Bare IP HTTP is not a supported Production edge—use a real domain. Details: requirements.

OS

Ubuntu 22.04 / 24.04 / 26.04 LTS amd64. Root or sudo required. arm64 is not supported.

CPU / RAM

Practical Production baseline: 4+ vCPU and 8+ GiB RAM. Lab/light minimums are lower. Add headroom when Stages run concurrently. Use the calculator above for concurrent-user planning.

Disk

SSD with free space for images, volumes, filestore, backups, and updates. Plan headroom before large backup or restore operations.

Network

Public ingress on 80 / 443 and SSH on 22. Keep application and database ports private behind Nginx.

Runtime

Docker Engine, Nginx, and certificate tooling are prepared by Soviez.sh. Soviez.sh never installs Webmin or Virtualmin.

Capacity map by concurrent operators (pair with the planner above). After install, preview sizing with soviez.sh --tune --dry-run.

Run once per server. The public bootstrap installs soviez.sh on PATH (typically /usr/local/bin/soviez.sh). Then prepare the host with --init before creating Production. Full install guide: Quick Start.

curl -fsSL https://soviez.sh | sudo bash

1. Installs the Soviez.sh launcher on PATH.
2. No GitHub account and no permanent Docker Hub login required.
3. sudo soviez.sh --init prepares Docker, Nginx, firewall, and the security baseline.
4. Never installs Webmin/Virtualmin; never opens port 10000 for Soviez.

From a clean supported Ubuntu host to Production: install Soviez.sh, prepare the host, point DNS at the server, then run --new. Guide: quick start.

curl -fsSL https://soviez.sh | sudo bash

# Prepare the host
sudo soviez.sh --init

# Create Production (domain, HTTPS, activation)
sudo soviez.sh --new

# Inspect
soviez.sh --list
soviez.sh --doctor

1. Pipe the public installer from https://soviez.sh.
2. Run --init to prepare the host.
3. Point DNS (for example erp.example.com) at the public IP, then run --new.
4. Open HTTPS on port 443. Keep application and database ports private.

Use soviez.sh for host prep, Production, updates, Stage, backup, restore, and diagnostics. Migration uses the --migration-* family only.

# Inventory & health
soviez.sh --version
soviez.sh --list
soviez.sh --doctor
soviez.sh --operations

# Resource sizing (preview first)
soviez.sh --tune --dry-run
soviez.sh --tune --explain

# Backup
soviez.sh --backup <production-id>
soviez.sh --backup-list

# Update (approved release)
soviez.sh --update <production-id> --release <release-id>

# Stage (when entitled)
soviez.sh --stage --production-tenant <id> --stage-domain stage.example.com
soviez.sh --stage-list

Install

--init prepares the host; --new creates Production with domain, HTTPS, and activation.

Inspect

--list, --doctor, --operations, and --security-status.

Update

--update applies an approved ERP release—not a mutable latest tag. See updates.

Stage is an isolated copy of Production for testing changes before they reach the live system. It uses a separate runtime, database, filestore, and hostname. Default lifetime is 14 calendar days, with a maximum of 60 days from original creation. Stage entitlement expiry blocks new Stage lifecycle operations—it does not automatically delete an existing Stage. Guide: Stage.

1. Confirm Stage entitlement on your Soviez account.
2. Create with a Production tenant id and a dedicated Stage hostname (--stage-domain).
3. Extend retention within the 60-day ceiling when needed.
4. Before automated cleanup, Soviez.sh protects recoverability (final backup and operator confirmation paths). Drop is operator-initiated.

# Create Stage from Production
soviez.sh --stage --production-tenant <id> --stage-domain stage.example.com

# Inventory & retention
soviez.sh --stage-list
soviez.sh --stage-retention-status
soviez.sh --stage-retention-extend <stage-id> --days N

# Operator teardown
soviez.sh --stage-drop <stage-id>

Inspect installation health with local commands. If an operation cannot complete safely, Soviez.sh stops and reports the problem rather than forcing an unknown state. Full recovery guide: recovery.

soviez.sh --list
soviez.sh --doctor
soviez.sh --operations
soviez.sh --security-status
soviez.sh --ssl-status

List

Inventory of managed Production and Stage environments.

Doctor

Read-only host and platform checks (OS, Docker, platform payload, firewall, AppArmor, Fail2Ban, resources).

Operations

Track long-running install/update/restore jobs; reattach after disconnect instead of starting duplicates.

Soviez.sh applies a production-oriented security baseline—not a claim of absolute security. After --new, open your domain over HTTPS, sign in with the printed admin credentials, and complete activation if prompted. Guide: security.

Edge

Public 22 / 80 / 443. ERP and PostgreSQL stay private behind Nginx.

Runtime

Non-root ERP containers, container isolation, least-privilege database access for the application role.

Host

Firewall defaults, Fail2Ban where armed, AppArmor expected enabled, unattended security updates where configured.

Malware

Scanning and quarantine workflows can flag suspicious content without automatically deleting business files.

Panels

Soviez.sh never installs Webmin or Virtualmin.

Connected operations such as device authorization, new Production, product updates, and new Stage operations may require reachability to Soviez services.

Independence: an existing ERP installation continues operating if Soviez is temporarily unreachable. Support or update entitlement expiry does not stop ERP, block login, or delete business data—only covered support and future update access are affected. Local backup, restore, status, doctor, and recovery remain available. Details: FAQ.

Backups cover database, filestore, and required configuration/state. Restore returns an environment from a backup. Prefer off-host copies for stronger disaster-recovery posture. Guides: backup · restore.

soviez.sh --backup <production-id>
soviez.sh --backup-list
soviez.sh --backup-verify <backup-id>
soviez.sh --restore <production-id> --backup <backup-id> --confirm

Move between Soviez hosts with the --migration-* command family (authorization, pairing, transfer, cutover, rollback). There is no single-flag migrate-in shortcut. Source cleanup is never automatic. Guide: migration.

Soviez.sh securely retrieves approved Soviez ERP releases for your licensed installation. You do not need permanent Docker registry credentials on the server. Install and update use exact approved releases—not a mutable latest tag.

1. You start install or update.
2. Soviez.sh verifies license access and selects the approved release.
3. The runtime is downloaded with short-lived, pull-only access, then temporary credentials are discarded.

More detail: private image delivery.

Full operator documentation, FAQ, and troubleshooting live on soviez.com/docs. For commercial questions see pricing. For help: support.