What we back up
- Postgres — your workflows and execution history
- The credentials store
- The
N8N_ENCRYPTION_KEY, stored separately - Configuration and environment
- Custom nodes and binary data
Security and backups for managed n8n hosting
Security and backups are the reason to pay for managed hosting. In late 2025 and early 2026 n8n disclosed multiple critical, unauthenticated remote-code-execution vulnerabilities — several at the highest severity possible, at least one added to the US government's known-exploited list — with tens of thousands of unpatched instances sitting exposed online. Most self-hosters never patch. We give you an isolated, backed-up instance and the infrastructure to patch safely on your own schedule — your n8n version and security patches stay in your control.
Isolated per-customer instanceRestores tested monthly30-day money-back guarantee
How your infrastructure supports safe patching
You control when your n8n version is updated. We recommend watching n8n's own release notes and security advisories and updating promptly. What we provide makes that safe to do.
1
Isolated instance
Your instance is single-tenant, with its own database and encryption key, so an update on your instance never touches anyone else's.
2
Backed up first
Daily off-site backups, tested monthly, mean an update that goes wrong is recoverable.
3
Behind Cloudflare
Edge protection and secure defaults reduce your exposure while you decide your own update schedule.
4
Infra maintenance, announced
Server and OS maintenance windows are announced; on Scale they are coordinated with you. Your n8n version and its updates remain yours to schedule.
Isolation
One instance per customer. A separate Postgres database, separate credentials and a separate N8N_ENCRYPTION_KEY stored apart from the database it decrypts. Container CPU and memory caps, network egress limits and per-instance webhook URLs.
Nothing is shared between customers — not the process, not the database, not the key.
Edge protection and secure defaults
- Cloudflare proxy in front of every instance — WAF and DDoS mitigation
- Public unauthenticated form and webhook endpoints are off by default
- Restricted OS privileges on the n8n process
- Higher-risk nodes disabled unless you ask for them
Backups and recovery
Daily, encrypted, off site — and restored on a schedule so we know they work.
Schedule, retention and recovery
- Daily backups on every plan
- Starter: 7-day retention. Business and above: 30-day, off-site, encrypted at rest
- Restores tested every month
- With daily backups you could lose up to ~24h of changes (RPO)
- Restore target 2–4 hours by plan (RTO)
Monitoring
- Uptime and health checks
- CPU, memory and disk
- Postgres health and the n8n process
- Failed-workflow rate and webhook errors
- SSL expiry and backup success
- Your current n8n version, visible to you — updating it is your call
Data portability
The honest answer to "what if you disappear?"
Request a full export at any time, or on cancellation: your database plus your workflow JSON. Take it to another host and run it. There is no lock-in beyond the ordinary work of switching providers.
Shared responsibility
| Outcome | We secure | You secure |
|---|---|---|
| Server, OS, Docker and network | Yes | — |
| n8n version and security patches | — | You |
| Troubleshooting inside your workflows | — | You |
| Backups, encryption at rest and restores | Yes | — |
| Edge protection and secure defaults | Yes | — |
| Your workflow logic and what it does | — | You |
| Your third-party API keys and credentials | — | You |
| Your login hygiene and who you share access with | — | You |
Service-level targets
Uptime figures below are targets, not guarantees. The Business column is shaded as the most common production choice.
| Outcome | Starter | Business | Agency | Scale |
|---|---|---|---|---|
| Uptime target | 99.5% | 99.7% | 99.9% | 99.9% |
| Support response target | 24–48h | Next business day | 4–8h | 4h |
| Incident / emergency response | Best effort | Emergency line | Priority + emergency line | Priority, 4h |
| Backup retention | 7 days | 30 days | 30 days | 30–90 days |
| Restore target (RTO) | < 4h | < 4h | < 2–4h | < 2h |
| Maintenance windows | Announced, off-peak | Announced, off-peak | Announced, off-peak | Announced + coordinated |
Exclusions: scheduled and announced maintenance, customer misconfiguration, third-party API or DNS/registrar outages, force majeure, and instances suspended for abuse. Full terms are in the Terms.
Compliance and data location
You choose whether your instance runs in the US or the EU. We operate on a GDPR basis and a Data Processing Agreement is available — see the DPA and the Privacy Policy, including the sub-processor list.
We hold no formal certification such as SOC 2 or ISO 27001 yet, and we don't claim one.
Security & backup questions
Who updates and patches my n8n instance?
You do. We recommend watching n8n's own release notes and security advisories and applying updates on your own schedule. We provide the isolated, backed-up infrastructure - including Cloudflare edge protection - that makes doing this safely straightforward.
Are backups included - how often, how long, and where?
Yes. Backups run daily. Retention is 7 days on Starter and 30 days on Business and above, stored off-site and encrypted at rest.
Do you test restores?
Yes, every month. An untested backup is not a backup.
Is my instance isolated from other customers?
Yes. Every customer gets a separate instance, database and encryption key, each with its own CPU and memory limits. Nothing is shared between customers.
Put your n8n on infrastructure built to be secured
Dedicated, isolated, monitored, and backed up with tested restores — with a human on support for your infrastructure. Standard migration is free.
30-day money-back guaranteeFree migrationCancel anytime