Hello Traefik community,
I'm running ERPNext and N8N on a Hostinger VPS with Traefik as a reverse proxy (all containerized with Docker). I'm looking to improve the resilience of my setup to minimize potential downtime.
Current Infrastructure:
Hostinger VPS
ERPNext deployed via Docker containers
Traefik as reverse proxy
N8N workflow automation tool
I've researched several approaches to enhance uptime and would appreciate feedback or corrections:
- Database separation: Move the ERPNext database to a separate remote server
- Monitoring solution: Implement real-time monitoring with ERPNext-CrateDB + N8N for alerts
- Failover system: Set up DNS failover to route traffic to a backup ERPNext instance (sharing the same database)
What would you recommend for maximizing uptime in this setup especially for Traefik configuration?
Thank you in advance.
This does not seem like a Traefik problem.
You need to ensure availability of domain and initial target IP. Multiple IPs on domain do not help, as client/browser will pick a random one, will not try another if the first fails.
Traefik needs to be redundant or in a VM that will automatically be re-created with all required dependencies, like IP and Traefik configuration.
Database should be a cluster or primary/replica, each with own storage. Usually you would not let a database write to a shared storage and let a backup take over the same storage.
We use a managed load-balancer with fixed IP, multiple Traefik servers, multiple application servers, database cluster with multiple servers.
We had VMs fail with their local storage, we had dedicated servers fail because of fan, power, even with RAID. Make sure you have multiple copies of your data for worst case.
