I’m no longer able to login to the Hivemq edge admin console after upgrading from 2024.9 to 2025.2. Hivemq edge is running in a docker container, no errors are logged, however chrome debug console reports GET http://127.0.0.1:8080/api/v1/frontend/configuration net::ERR_CONNECTION_REFUSED index-C4zGATjt.js:441. This error is repeated multiple times and Network Error We cannot load your default credentials. Please try again later is displayed above the user name text box.
Hivemq edge is running and data bridges are working as expected, I’m just unable to login to the admin. I tried adding the URL to /etc/hosts in the docker container without success.
This is hosted behind a nginx reveres proxy, so the api calls should be to https://hivemq-edge.domain.com.
I’m assuming something in the config changed? The release notes on github take me to a 404 page, so maybe I’m missing a breaking change?
Hi @PrairieSnpr
It looks like there might be a port conflict between HiveMQ Edge and Nginx. Could you try stopping Nginx temporarily and then accessing the HiveMQ Edge admin console directly at http://127.0.0.1:8080? This will help determine if Nginx is interfering with the connection or if the issue lies elsewhere.
Let us know what you find
Best,
Dasha from The HiveMQ Team
Thanks @Daria_H, we can rule out an Nginx conflict, the web ui is visible, so that portion is working. I can’t really test as you asked since this is a headless server, but the web ui shows as expected. The issue is, that Nginx correctly proxies https://hivemq-edge.domain.com to the docker container on port 8080, but the script that runs in the browser is still trying to reach localhost on port 8080. I assume that some logic was moved from the server to the client and I need to set the base url, but I have yet to find a setting for that.
Another note, reverting to 2025.1 fixes this, the issue is isolated to 2025.1-k8s.
Hi PrairieSnpr,
It sounds like you’re not encountering issues with the hivemq-edge:2025.1 Docker image, but you’re having trouble with the hivemq-edge:2025.1-k8s image.
You mentioned that the image is running in a Docker container, but the -k8s tag is typically meant for Kubernetes environments. Could you clarify why you’re running the Kubernetes-specific image in a Docker container?
Looking forward to your response!
Best,
Dasha from The HiveMQ Team
I actually was using the latest tag. Looking at https://hub.docker.com/r/hivemq/hivemq-edge/tags, the sha256 hash is the same for latest and 2025.1-k8s.
This is resolved in 2025.2.