How to change DNS cache TTL in n8n-docker (or flush cache)?

Questions
1

Describe the problem/error/question

Every second attempt to test credentials for my gitlab-node failed inside my n8n-docker-container.

This is a DNS-issue, because a ping inside the docker-container also resolves to different IP-adresses.

ping gitlab.local
PING gitlab.local (192.168.1.166)
ping gitlab.local
PING gitlab.local (192.168.1.23)
ping gitlab.local
PING gitlab.local (192.168.1.166)
...

…you get the idea.
BUT inside my DNS-server there is right now only one domain IP-Adress-Record
And all my other docker-containers / computers get only the right IP (192.168.1.166) for gitlab.
And n8n gets still the old AND the new IP-adress….even, if i restart the hole system.

Most solutions how to solve this issue from the internet are not working inside the docker-container, because the tools are not available in unmodified n8n-container.

Does anyone here know a working solution for me?
Thanks!

2 
# docker-compose.yml - Add DNS configuration to your n8n service

services:
  n8n:
    image: n8nio/n8n
    # ... your existing config ...
    
    # Option 1: Specify DNS servers directly (recommended)
    dns:
      - 192.168.1.1      # Your local DNS server
      - 8.8.8.8          # Google DNS fallback
    
    # Option 2: Add static host entry (if IP is fixed)
    extra_hosts:
      - "gitlab.local:192.168.1.166"

# Flush host DNS cache (run on Docker host)

# Ubuntu/Debian with systemd-resolved
sudo systemd-resolve --flush-caches
# or
sudo resolvectl flush-caches

# Restart Docker daemon
sudo systemctl restart docker

# Recreate container (not just restart)
docker-compose down && docker-compose up -d

// /etc/docker/daemon.json - Global Docker DNS config (alternative)
{
  "dns": ["192.168.1.1", "8.8.8.8"]
}

# After editing daemon.json
sudo systemctl restart docker
docker-compose down && docker-compose up -d
3

Hi, sorry for my late answer,

but one comment helped me to guide me in the right direction: “Flush host DNS cache (run on Docker host)”

The issue was not in n8n-container or in my DNS-server or in the connection between them.
The rootcause was a undocumented DNS-cache on Docker-Host.
On my NAS there where no possibility to flush this cache (to low level).
But after i use a different virtual switch/gateway between docker-container and network-adapter the issue was gone. :slight_smile: