Compare, self-hosted

Levelrail vs Dokku

The classic Heroku on your own box: git push to deploy, buildpacks, a rich plugin ecosystem.

As of 2026-10. Statements about Dokku come from its own public documentation, linked on each row.

The short verdict

Dokku is good at: almost no idle footprint, because there is no long-running control service; and hard to beat for simplicity on a single server.

Choose Dokku when you run one server and want the simplest possible thing.

Levelrail differs here: a dashboard with deploy history and one-click rollback to a pinned previous image; and metrics and log search in the core, where Dokku suggests adding your own stack; and more servers join with a one-time token and an agent that dials out. Levelrail is pre-release, a single node is the well-tested path and multi-node is in beta, so its track record is much shorter than Dokku's.

Where Dokku is ahead

No row where its docs say yes and ours do not. That reflects what its docs state, not maturity or the size of its user base.

Where Levelrail is ahead

  • No SSH: agents dial out. Levelrail: Agents dial out over mTLS; no inbound ports. Dokku: no, commands run on the host, over SSH or git push.
  • Level-triggered reconciler with status reasons. Levelrail: Desired state converged, status condition with reason per pass. Dokku: no, event-driven trigger pipeline; no reconcile loop described.
  • Metrics and log search built in. Levelrail: Node-local metrics at 15s and full-text log search. Dokku: no, add your own stack.
  • Crashloop diagnosis with last log lines. Levelrail: Crashloop detection surfaces the last log lines. Dokku: partly, restart policy restarts failing containers; no crashloop detection.
  • One-click rollback with pinned images. Levelrail: Prior images pinned so cleanup cannot remove them. Dokku: partly, restores last good deploy when a deploy fails; no manual rollback.
  • CLI and HTTP API. Levelrail: CLI and REST API documented; MCP server is beta. Dokku: partly, cLI over SSH; JSON API only in paid Dokku Pro.
  • IAM, audit log and 2FA in the free edition. Levelrail: Allow/Deny policies, audit log and TOTP in one edition. Dokku: partly, sSH key users; admin keyword grants key management; no RBAC.

28 capabilities side by side

Yes, partly and no are what each project's own documentation states. "Not stated" means their docs do not say, which is not the same as no. Levelrail maturity tags (beta, stable) come from its feature status page.

Levelrail compared with Dokku, by capability
Capability Levelrail Dokku
Architecture
No SSH: agents dial out Yesbeta Agents dial out over mTLS; no inbound portsSource for Levelrail, No SSH: agents dial out No Commands run on the host, over SSH or git pushSource for Dokku, No SSH: agents dial out
Level-triggered reconciler with status reasons Yes Desired state converged, status condition with reason per passSource for Levelrail, Level-triggered reconciler with status reasons No Event-driven trigger pipeline; no reconcile loop describedSource for Dokku, Level-triggered reconciler with status reasons
Single control plane binary Yes One control plane binary, plus one agent per extra nodeSource for Levelrail, Single control plane binary Not applicable
Observability
Metrics and log search built in Yes Node-local metrics at 15s and full-text log searchSource for Levelrail, Metrics and log search built in No Add your own stackSource for Dokku, Metrics and log search built in
Crashloop diagnosis with last log lines Yes Crashloop detection surfaces the last log linesSource for Levelrail, Crashloop diagnosis with last log lines Partly Restart policy restarts failing containers; no crashloop detectionSource for Dokku, Crashloop diagnosis with last log lines
Deploys
One-click rollback with pinned images Yes Prior images pinned so cleanup cannot remove themSource for Levelrail, One-click rollback with pinned images Partly Restores last good deploy when a deploy fails; no manual rollbackSource for Dokku, One-click rollback with pinned images
Readiness-gated deploys (zero-downtime on one node) Yes Cutover only after the readiness probe passes. Live-tested on one node; multi-node is betaSource for Levelrail, Readiness-gated deploys (zero-downtime on one node) Yes New containers start, proxy switches, old retired after delaySource for Dokku, Readiness-gated deploys (zero-downtime on one node)
Preview environments per pull request Yesstable Stable on GitHub, beta on GitLab and BitbucketSource for Levelrail, Preview environments per pull request Not stated in their docs
Docker Compose apps Partly Pre-built images from the dashboard; git deploys can buildSource for Levelrail, Docker Compose apps Not stated in their docs
Git push deploys from GitHub, GitLab, Bitbucket Partly All four documented; GitLab and Bitbucket previews are beta, synthetic-webhook testedSource for Levelrail, Git push deploys from GitHub, GitLab, Bitbucket Partly git push to deploy branch; git:sync from any cloneable repo; no webhookSource for Dokku, Git push deploys from GitHub, GitLab, Bitbucket
Builds from source without a Dockerfile Partly Railpack detects Node, Python, Go and more; Ruby is rejectedSource for Levelrail, Builds from source without a Dockerfile Partly Heroku buildpacks by default; pack and Railpack need config filesSource for Dokku, Builds from source without a Dockerfile
Scheduled jobs Yes Per-app cron jobs with run history, logs, retries and alertsSource for Levelrail, Scheduled jobs Yes Scheduled dokku run tasks via cron:set and app.jsonSource for Dokku, Scheduled jobs
Data
Managed databases with scheduled backups Partly Many engines, scheduled S3 backups; Postgres restores tested liveSource for Levelrail, Managed databases with scheduled backups Partly Official plugins create databases; backups manual export, no schedulingSource for Dokku, Managed databases with scheduled backups
Encrypted secrets for env vars Yes Envelope-encrypted, masked in UI, injected at container createSource for Levelrail, Encrypted secrets for env vars Not stated in their docs
Networking
Private networking between apps or nodes Partly Named networks and allow rules; cross-server mesh unverified on real hostsSource for Levelrail, Private networking between apps or nodes Partly Docker networks on one host only; no cross-host networkingSource for Dokku, Private networking between apps or nodes
Load balancing across replicas Partly Experimental and flag gated; not tried across real hostsSource for Levelrail, Load balancing across replicas Partly Nginx round-robin across scaled web containers; other processes need configSource for Dokku, Load balancing across replicas
Request firewall, rate limits, bot protection Yes Ordered rules, IP bans, bot handling, challenges and rate limitsSource for Levelrail, Request firewall, rate limits, bot protection Not stated in their docs
Automatic TLS certificates Partly Internal issuer by default; public ACME verified once liveSource for Levelrail, Automatic TLS certificates Partly Let's Encrypt via official plugin; manual certs otherwiseSource for Dokku, Automatic TLS certificates
Platform
Multi-node Partlybeta Enrollment verified locally; WireGuard mesh still unverifiedSource for Levelrail, Multi-node Partly One host by default; other schedulers existSource for Dokku, Multi-node
One-click service templates Yesbeta About 360 templates; a sample is live-deployed in e2eSource for Levelrail, One-click service templates Not stated in their docs
Autoscaling Partly Beta: CPU, memory or request-rate scaling with cooldowns and reasonsSource for Levelrail, Autoscaling No Scaling is manual via ps:scale; no autoscalingSource for Dokku, Autoscaling
CLI and HTTP API Yes CLI and REST API documented; MCP server is betaSource for Levelrail, CLI and HTTP API Partly CLI over SSH; JSON API only in paid Dokku ProSource for Dokku, CLI and HTTP API
Deploy and alert notifications Partly 18 channel kinds in beta, tested only against mocksSource for Levelrail, Deploy and alert notifications Partly Deploy webhook via community plugin; no built-in notificationsSource for Dokku, Deploy and alert notifications
MCP / AI assistant integration Yesbeta MCP server with over 240 tools in on-demand toolsetsSource for Levelrail, MCP / AI assistant integration Not stated in their docs
Runs on servers you own Yes Runs on Linux servers you provideSource for Levelrail, Runs on servers you own Yes Scripts running on your own hostSource for Dokku, Runs on servers you own
Open source license Yes Apache 2.0, one edition, no paid tierSource for Levelrail, Open source license Yes Permissive MIT-style license per repo LICENSE fileSource for Dokku, Open source license
Security
IAM, audit log and 2FA in the free edition Yesstable Allow/Deny policies, audit log and TOTP in one editionSource for Levelrail, IAM, audit log and 2FA in the free edition Partly SSH key users; admin keyword grants key management; no RBACSource for Dokku, IAM, audit log and 2FA in the free edition
SAML or SCIM single sign-on Partly SAML and SCIM are experimental and off by defaultSource for Levelrail, SAML or SCIM single sign-on Not stated in their docs

All ten projects are on the full matrix. Product names and logos belong to their owners and are used for identification only.

Migrating from Dokku

No importer for Dokku. Deploy the same repositories and recreate environment variables and domains.

Dokku alternative guide

Longer Dokku comparison in the docs

Questions about Levelrail and Dokku

Is Levelrail a good alternative to Dokku?

The classic Heroku on your own box: git push to deploy, buildpacks, a rich plugin ecosystem. How Levelrail differs: A dashboard with deploy history and one-click rollback to a pinned previous image. Metrics and log search in the core, where Dokku suggests adding your own stack. More servers join with a one-time token and an agent that dials out. Choose Dokku instead when you run one server and want the simplest possible thing. Levelrail is pre-release, so check the comparison for what it does not do yet.

When should I choose Dokku instead of Levelrail?

Choose Dokku when you run one server and want the simplest possible thing.

Can I move from Dokku to Levelrail?

No importer for Dokku. Deploy the same repositories and recreate environment variables and domains.

Is Levelrail ready to replace Dokku in production?

Pre-release: no stable release; APIs, the app spec and the on-disk layout can change between betas. A single node is the well-tested path; multi-node is in beta and the WireGuard mesh is not battle-tested. Check the feature status page before moving anything you cannot afford to lose, and keep Dokku running alongside until you trust the result.

Other comparisons