Fix: n8n container running but not working

    When Podman says n8n is running but localhost:5678 cannot connect, the failure is usually port publishing, container state, or binding—not the n8n migrations.

    Last updated 2026-09-21~5 min read
    n8nPodmanDocker-compatible containersLinux

    Last updated: 2026-09-03 · Tested on: n8n 1.x Docker image with Podman · ~5 min read

    Symptoms — what you're seeing

    • Podman reports that the n8n-server container is running, but http://localhost:5678 shows “This site can’t be reached”
    • The container log says n8n ready on ::, port 5678
    • Database migrations finish, or continue for a long time, without an n8n editor page opening
    • Reinstalling Podman and pulling the n8n image again does not change the result
    • There is no obvious error in the container log

    Why this happens

    The log line n8n ready on ::, port 5678 only confirms that n8n is listening inside the container. It does not confirm that the host published port 5678 or that your browser is connecting to the machine running Podman.

    The most common causes are:

    1. The container is not currently running. The logs can show a successful startup from an earlier run. Check the current state with podman ps -a, not only with an old log output.

    2. Port 5678 was not published. The -p 5678:5678 option is what connects the host port to n8n's container port. Without that mapping, n8n can be healthy inside the container and still be unreachable from the host.

    3. You are using localhost from the wrong machine. If Podman runs on a VPS, virtual machine, WSL environment, or another computer, localhost means the machine running your browser. It does not mean the remote host running n8n.

    4. The host mapping is limited to loopback. A mapping such as 127.0.0.1:5678->5678/tcp works from the Podman host but not from another device. A remote browser needs the server's address and a firewall rule that permits the connection.

    5. The process is listening on an address that does not match the published path. The :: in the log is an IPv6 wildcard address. Podman normally handles this correctly, but explicitly setting n8n to listen on 0.0.0.0 removes the ambiguity when IPv4 access is failing.

    The fix, step by step

    1. Check the current container state. Run this on the same machine where Podman is running:
      podman ps -a --filter name=n8n-server
      The status should be Up or Running. If it is Exited, inspect the current failure instead of relying on the long migration output:
      podman logs --tail 100 n8n-server
      podman inspect n8n-server --format '{{.State.Status}} {{.State.ExitCode}}'
    2. Verify the published port. Run:
      podman port n8n-server
      You need to see a host mapping for container port 5678. A result such as 5678/tcp -> 0.0.0.0:5678 means Podman published it on all host interfaces. If the command returns nothing, the existing container was created without a usable port mapping.
    3. Test the port without the browser. From the Podman host, run:
      curl -i http://127.0.0.1:5678/
      An HTTP response proves that the host can reach n8n. The response may be a redirect or HTML rather than a JSON API response; that is enough for this connectivity test. A connection refusal means the container is stopped, the port is not published, or the host listener is unavailable.

    Still blocked after the host-side test? Book a 30-min scope callEmail Basit

    1. Recreate the container with an explicit mapping if the port is missing. The named volume preserves n8n's data, including the encryption key and workflows, so remove only the container—not the n8n_storage volume:
      podman rm -f n8n-server
      podman pull docker.io/n8nio/n8n:latest
      podman run -d   --name n8n-server   --restart unless-stopped   -p 5678:5678   -e N8N_LISTEN_ADDRESS=0.0.0.0   -v n8n_storage:/home/node/.n8n   docker.io/n8nio/n8n:latest
      Do not run podman volume rm n8n_storage. That deletes the persisted n8n configuration and workflows.
    2. Open the correct address. If the browser is on the same machine as Podman, use http://127.0.0.1:5678. If Podman runs on another host, replace 127.0.0.1 with that host's reachable IP or DNS name. Check the mapping first; do not expose n8n publicly until authentication, TLS, and network access are configured.
    3. Check the host firewall only for remote access. If curl http://127.0.0.1:5678/ works on the server but a different machine cannot connect, the n8n container is working. Check the server firewall, cloud security group, and whether Podman published the port on the required interface — exact firewall commands vary by host distribution, so confirm the right one for your server before running it.

    Common variations

    • “localhost:5678 works on the server but not on my laptop” — localhost is resolving to the laptop. Use the server address, and verify the published port and firewall path.
    • “The container is Up but the browser still refuses the connection” — run podman port n8n-server and curl -i http://127.0.0.1:5678/. The container state alone does not prove that the host port is reachable.
    • “The log stops during migrations” — inspect the last log lines and container exit code. Do not delete the volume before checking the error; the volume contains the n8n database and encryption key.
    • “The page loads but scheduled workflows run at the wrong time” — that is a separate timezone configuration issue. See the n8n schedule trigger wrong timezone fix.

    When this isn't the issue

    If the host-side curl request returns an HTTP response but the browser still fails, the container and port mapping are not the problem. Continue with these checks:

    • Confirm that the browser URL is http://, not https://, unless you configured a reverse proxy with TLS
    • Check whether a proxy, VPN, browser extension, or corporate network is blocking the address
    • If you are using a domain or reverse proxy, verify its upstream points to the correct host port and inspect the proxy logs
    • If the page opens but a workflow fails after startup, diagnose the workflow node and credentials separately; connectivity is no longer the failure

    Need help with the container?

    If this is still broken after the container is running, the port is published, and the host-side curl test succeeds, I'll get on Zoom, inspect the Podman state and n8n logs, and either fix it on the call or scope what it will take. No upsell, no decks.

    Book a 30-min scope callEmail Basit

    Related pages

    Need this workflow customized or debugged? Book a 30-min scope call →

    Stuck configuring this workflow?

    I build and fix n8n workflows for agencies and SaaS teams. 30-minute scope call — bring the problem, leave with an estimate.

    Related pages