Frequently Asked Questions¶

Common questions about m9m.

General¶

What is m9m?¶

m9m is a high-performance workflow automation platform written in Go. It provides n8n-compatible workflow execution with significantly better performance characteristics.

How does m9m compare to n8n?¶

Feature	m9m	n8n
Execution speed	5-10x faster	Baseline
Memory usage	~150 MB	~512 MB
Container size	~300 MB	~1.2 GB
Startup time	~500 ms	~3 s
Workflow format	Compatible	Original
Language	Go	TypeScript

Is m9m compatible with n8n workflows?¶

Yes. m9m can execute n8n workflows without modification. Simply export your workflow from n8n and run it with m9m.

Is m9m open source?¶

Yes, m9m is open source and available on GitHub.

Installation¶

What are the system requirements?¶

Minimum: - 1 CPU core - 256 MB RAM - 1 GB disk

Recommended: - 2+ CPU cores - 512 MB+ RAM - 10 GB+ SSD

Which databases are supported?¶

SQLite (default, single instance)
PostgreSQL (production, scalable)

Can I run m9m on Windows?¶

m9m is primarily tested on Linux. For Windows, we recommend using Docker or WSL2.

Workflows¶

How do I create a workflow?¶

Three options:

JSON file - Write workflow JSON directly
API - Create via REST API
n8n export - Export from n8n and import

Can I use n8n's web editor?¶

m9m focuses on headless execution. You can: - Use n8n's editor to design workflows, then export - Use the API to manage workflows programmatically - Write JSON workflows directly

How many workflows can I run?¶

There's no hard limit. Performance depends on: - Hardware resources - Workflow complexity - Execution frequency

What node types are supported?¶

34+ node types including: - HTTP requests - Database operations (PostgreSQL, MySQL, SQLite) - Messaging (Slack, Discord) - AI/LLM (OpenAI, Anthropic) - Cloud services (AWS, Azure, GCP) - And more

See Nodes documentation for full list.

Execution¶

How do I trigger a workflow?¶

Manual - CLI or API call
Webhook - HTTP request to webhook URL
Schedule - Cron expression

What happens if a workflow fails?¶

Execution is marked as "failed"
Error details are logged
Configured retries are attempted
Alerting can be configured

Can workflows run in parallel?¶

Yes. Multiple workflows can execute simultaneously. Configure the number of workers:

queue:
  workers: 10

How long can a workflow run?¶

Default timeout is 5 minutes. Configure per-workflow:

execution:
  timeout: 30m

Performance¶

Why is m9m faster than n8n?¶

Go vs JavaScript - Compiled language, better performance
Efficient memory - Less garbage collection overhead
Optimized execution - Parallel node execution where possible
Native concurrency - Goroutines for efficient parallelism

How do I optimize workflow performance?¶

Minimize external API calls
Use caching where possible
Process data in batches
Use async execution for long workflows
Monitor and profile executions

What's the maximum data size?¶

There's no hard limit, but large datasets may impact performance. For large data: - Use streaming where possible - Process in batches - Increase memory allocation

Security¶

How are credentials stored?¶

Credentials are encrypted at rest using AES-256. The encryption key should be stored securely (environment variable, secrets manager).

Does m9m support SSO?¶

Currently, m9m supports JWT and API key authentication. SSO integration is on the roadmap.

Is data encrypted in transit?¶

Configure TLS for HTTPS:

server:
  tls:
    enabled: true
    certFile: /path/to/cert.pem
    keyFile: /path/to/key.pem

Deployment¶

Should I use Docker or Kubernetes?¶

Use Case	Recommendation
Development	Docker
Single server	Docker Compose
Production	Binary / package manager
Edge deployment	Binary
Kubernetes	Experimental

How do I scale m9m?¶

Run multiple instances behind load balancer
Use Redis for distributed queue
Use PostgreSQL for shared storage
Use orchestration only if you are prepared to run experimental Kubernetes manifests

How do I backup m9m?¶

Backup the database:

# SQLite
cp /data/m9m.db /backup/m9m-$(date +%Y%m%d).db

# PostgreSQL
pg_dump m9m > /backup/m9m-$(date +%Y%m%d).sql

Troubleshooting¶

Where are the logs?¶

# CLI
m9m serve  # Outputs to stdout

# Docker
docker logs m9m

# Kubernetes
kubectl logs deployment/m9m

Why is my webhook not working?¶

Is the workflow activated?
Is the path correct?
Is the HTTP method correct?
Check server logs for errors

Why is my execution stuck?¶

Check for: - Long-running external calls - Infinite loops in Code nodes - Resource exhaustion

Cancel stuck execution:

m9m execution cancel exec-123

Migration¶

How do I migrate from n8n?¶

Export workflows from n8n (JSON format)
Import into m9m: m9m create --from workflow.json
Create credentials in m9m
Test workflows
Activate workflows

Are all n8n nodes supported?¶

Most common nodes are supported. Check the nodes documentation for the full list.

Can I run n8n and m9m together?¶

Yes, you can run both during migration: - Keep n8n for unsupported workflows - Migrate supported workflows to m9m - Gradually transition

Support¶

How do I report a bug?¶

Open an issue on GitHub with: - m9m version - Steps to reproduce - Expected vs actual behavior - Relevant logs

How do I request a feature?¶

Open a feature request on GitHub describing: - Use case - Proposed solution - Alternatives considered

Is commercial support available?¶

Contact Neul Labs for commercial support options.