Self-hosted n8n in queue mode, on a managed database layer
n8n in queue mode, with PostgreSQL and Redis or Valkey underneath it, run for you in your environment.
45 minutes with an engineer: we review your workload and sketch the design.
- ✓Operated by AceMQ: 160+ customers, 10+ years
- ✓24/7 on-call engineers
Self-hosted n8n in queue mode splits work between a main instance, worker processes and a Redis queue, with a database that must not be SQLite. OSStend runs n8n on managed PostgreSQL and Redis or Valkey, your choice, in your environment. n8n Cloud's published plans start at €20 a month billed annually, by executions.
- PostgreSQL
- MySQL
- MongoDB
- Redis
- Valkey
- OpenSearch
- Kafka
- RabbitMQ
- AWS
- Google Cloud
- Azure
- Kubernetes
- OpenShift
- Prometheus
- Grafana
- Datadog
- Dynatrace
- Splunk
What is n8n queue mode?
Queue mode lets n8n run executions on separate worker processes instead of the main instance. n8n's docs describe a main instance that handles timers and webhook calls, Redis as the message broker, workers that run executions, and optional webhook processors.
| Part | What it does (per n8n) | With OSStend |
|---|---|---|
| Main instance | Handles timers and webhook calls and generates executions | Deployed and upgraded for you |
| Workers | Separate Node.js processes that run the executions | Sized to your execution load |
| Webhook processors (optional) | Handle incoming webhooks and scale that layer separately | Added where webhook load needs them |
| Redis | The queue of pending executions and the message broker | Managed Valkey or Redis. See managed Redis or Valkey |
| Database | Persists data; SQLite is not supported for a distributed queue-mode setup | Managed PostgreSQL. See managed PostgreSQL |
Source: n8n queue mode docs, checked 17 Nov 2026. n8n's docs name Redis; OSStend runs the queue on Redis or Valkey, whichever you prefer, and Valkey speaks the same protocol. OSStend runs the service in your environment, such as your cloud account or your data center.
There is no fixed menu here: the worker count, retention, versions and upgrade dates your workload needs are written into the design you sign off before anything is built, and that design is what OSStend runs.
What goes wrong when self-hosting n8n at scale?
Most problems come from the parts around n8n, not n8n itself: the database, the queue, binary data and keeping every process configured the same. n8n's docs name each of them.
- `EXECUTIONS_MODE=queue` must be set on the main instance and on every worker, and `N8N_ENCRYPTION_KEY` must match the main instance on every worker and webhook processor so credentials can be read (n8n docs, checked 17 Nov 2026).
- In queue mode, n8n stores binary data in the database by default. The docs say external storage such as AWS S3 or Azure Blob Storage depends on your license, and filesystem storage needs a shared volume at the same path on every process (n8n docs, checked 17 Nov 2026).
- A multi-main setup requires Postgres and Redis for all main processes (n8n docs, checked 17 Nov 2026).
- The database is where execution history accumulates, so retention, backups and growth need a plan.
From our on-call engineers: n8n queue mode settings that cause outages
n8n queue mode splits the editor from the workers, and most failures are misconfiguration between them. These are the settings we check first, taken from n8n's own documentation.
- ✓Queue mode needs a main instance, Redis as the broker, workers and a shared database, with optional webhook processors scaled separately behind a load balancer.
- ✓Workers and webhook processors must use the main instance's N8N_ENCRYPTION_KEY, or they cannot read credentials stored in the database, so the key is backed up with the database.
- ✓n8n says running a distributed queue-mode setup over SQLite is not supported, so queue mode means PostgreSQL.
- ✓Webhook routing sends /webhook/* and /webhook-waiting/* to the webhook pool and everything else to main, and the main process is not added to the webhook pool.
- ✓Pruning is on by default, removing finished executions older than 336 hours or beyond 10,000 total, so a busy instance that needs longer history has to change the limits deliberately and plan database growth.
What does n8n cost: Cloud or self-hosted?
n8n Cloud prices by monthly workflow executions, from €20 a month billed annually. The Community Edition is free to self-host, and n8n sells Business and Enterprise editions for self-hosting that need a license key.
| Plan | Price per month (billed annually) | Executions per month | Concurrent executions |
|---|---|---|---|
| Starter | €20 | 2,500 | 5 |
| Pro | €50 | 10,000 | Up to 50 |
| Business | €667 | 40,000 | Not listed |
| Enterprise | Custom | Custom | 200 or more |
Source: n8n pricing, checked 17 Nov 2026. n8n states that pricing is based on monthly workflow executions regardless of complexity. Business is self-hosted only (a hosted version is on a waitlist). The page's plan cards and comparison table disagree on whether SSO, environments, Git version control and queue mode sit in Business or Enterprise, so confirm with n8n before relying on either.
n8n describes its code as fair-code, under the Sustainable Use License and the n8n Enterprise License (GitHub, checked 17 Nov 2026). Read both against how you plan to use it before choosing an edition.
n8n Cloud, self-hosted yourself or managed by OSStend?
n8n Cloud is the quickest start, self-hosting keeps workflows and credentials in your environment, and OSStend runs the self-hosted stack so your team does not. Choose by where the data may live and who is on call.
| Factor | OSStend | n8n Cloud | Self-hosted yourself |
|---|---|---|---|
| Who runs it | OSStend engineers, 24/7 | n8n | Your team |
| Where workflows and credentials live | n8n and its databases in your environment, such as your cloud account or your data center | n8n Cloud | Wherever you deploy it |
| Pricing | Quoted per design. See pricing | €20, €50 or €667 a month by plan, billed annually (n8n pricing) | Community Edition is free; you pay for the infrastructure and the people |
| Database and queue | Managed PostgreSQL and Redis or Valkey | Run by n8n | You run both |
| Good fit | An enterprise team that wants n8n in its environment and run for it | A small team that wants to start quickly | A platform team already running PostgreSQL and Redis |
| Poor fit | A hobby project or a free tier | Workflows that cannot leave your environment | No one to carry the pager |
How we compared: n8n's pricing page and queue mode docs, checked 17 Nov 2026. OSStend's price is set after a workload review.
How does moving to managed n8n work?
The new n8n stack is built beside your current instance, then workflows and credentials move across once it is verified. The first step is a 45-minute workload review with an engineer: the current setup is reviewed, the design is sketched, and a price follows for that design.
Review
An engineer reviews the current setup: sizes, settings, modules or plugins, versions, traffic and where it must run.
Design
The cluster design is written and signed off before anything is built: size, topology, configs, versions and upgrade calendar.
Move
The new cluster runs beside the old one and stays current while the old service stays live. Traffic switches only when the new side is verified.
Run
Named engineers operate it 24/7: backups, patching, scaling and incident response, with upgrades on the customer's change calendar.
Questions about self-hosting n8n
These are the questions teams ask when they move n8n off a single server.
Does n8n need PostgreSQL in queue mode?
n8n's docs say a distributed queue-mode setup over SQLite is not supported, and a multi-main setup requires Postgres and Redis (n8n docs, checked 17 Nov 2026). OSStend runs n8n on managed PostgreSQL.
Can n8n use Valkey instead of Redis?
n8n's docs name Redis. OSStend runs the queue on Redis or Valkey, whichever you prefer, and Valkey speaks the same protocol.
How much does n8n cost?
n8n Cloud lists €20, €50 and €667 a month billed annually, by monthly executions; Community Edition is free to self-host (n8n pricing, checked 17 Nov 2026).
Can OSStend run n8n in my own cloud account?
Yes. OSStend runs n8n in your environment, such as your cloud account or your data center, to a design you sign off before anything is built.
When is n8n Cloud the better choice?
When you want the fastest start, a small team runs the workflows, and nothing needs to stay in your environment.
Related pages
- Redis or ValkeyManaged Redis or Valkey, your choice of engineManaged Redis or Valkey: you choose the engine, with persistence, eviction and configuration set for your workload, run 24/7 in your cloud or data center.
- PostgreSQLManaged PostgreSQL service and Postgres hostingManaged PostgreSQL service, compared: how to choose between RDS, Aiven, Crunchy Bridge, Neon and a tailored cluster run 24/7 in your cloud or ours.
- PricingDatabase as a service pricing: dedicated, tailoredDatabase as a service pricing from OSStend: dedicated, tailored clusters priced by engine, size, SLA tier and location. Enterprise only, with no free tier.
Talk to an engineer about n8n
45 minutes: bring your execution volume and where the workflows must run, and the engineer sketches the n8n stack.
- ✓A design sketch for your workload, not a plan menu
- ✓A zero-downtime cutover path for what you run today
- ✓A price for that design, after the review