We recently started using n8n under an enterprise license. Before that, we were running on a trial key. Now that we’re officially licensed, we updated the DNS for our machines to use more formal company URLs.
Our deployment is running in Kubernetes. After making the DNS changes, the machines came back up, but we’re seeing some odd behavior. We have a testing project set up with quite a few workflows in it. The workflows still exist, because if you go directly to a workflow URL, it opens correctly. But they are not showing up in the UI.
For example, the UI shows 17 workflows under the “Agents” project, but when you click into “Agents,” no workflows are displayed. See screenshot.
welcome to the n8n community @Skyler76
I’d start by opening n8n in an incognito window or doing a hard refresh to clear any cached data. If that doesn’t fix it, I’d restart the n8n pods so the backend and editor reload cleanly with the new domain. After that, I’d navigate back to the root project (“Corporate-Testing”) and enter the “Agents” folder again instead of relying on a bookmarked URL. That sequence has worked for me in similar DNS/domain changes and usually brings the workflow list back.
Hey! Check your env vars like N8NHOST and WEBHOOKURL, they probably still point to the old DNS. Update those in your ConfigMap to match the new URLs and restart the pods, should fix the UI.
@Skyler76
I did some research and found that this is most likely related to the new URL/proxy setup rather than DNS itself. This can happen when n8n is exposed behind a reverse proxy using N8N_PATH, which may break folder/project navigation even though direct workflow URLs still work. The fix is usually to move n8n to a subdomain or remove N8N_PATH, then restart all pods. If you’re running multiple pods, also make sure they all share the same Postgres/Redis and the same N8N_ENCRYPTION_KEY.
Since your env vars are good, this might be a browser/cache thing with the new domain, have you tried clearing your browser storage or using a different browser entirely? Also worth checking if theres anything in the n8n pod logs when you open that folder.
ran into this after k8s upgrades too. the workflow list cache isnt always cleared on pod restart when the db is external. we ended up checking the logs for ‘failed to list workflows’ errors, then manually cleared n8ns internal cache via the db (workflows table update). painful but fixed it. what db are you using?
postgres. Managed to get it resolved. We ran into some issues when we changed the domain and went through some troubleshooting steps to get the system back up. I think one of those steps reassigned the workflows to a user, rather than a project. We reassigned them back and now they’re there. Bit odd to say the least.
