posts/homepage-dashboard/

[giscus[bot]]

posts/homepage-dashboard/

Meet Homepage, your new HomeLab services dashboard homepage! Homepage is an open source, highly customizable homepage (or startpage) dashboard that runs on Docker and is integrated with over 100 API services. It’s easy to set up, looks good by default, and helps you keep track of everything you are running in your HomeLab and more. Today we’ll set up Homepage and get it running in Docker in no time.

https://technotim.live/posts/homepage-dashboard/

[dwilliss]

This is great but I can't get the Proxmox widget to show any data. It doesn't give an error, it just shows 0/0 VMS, 0/0 LXC NaN% CPU NaN% Mem.
I followed the instructions for how to generate a token for the user "api" and token name of "homepage". I configured the widget with the user name "api@pam!homepage" and the token that was generated.

All the other widgets I've tried work fine.

I even got the pfSense one to work after upgrading my pfSense to 2.7.2 and installing the api plugin that the homepage documentation says to use. Pretty scary because after the reboot, it took 15 minutes for pfSense to come back up and I thought I had borked it all up.

[brian1xevdo]

I had a difficult time with ProxMox too. There is a great guide on the homepage website: https://gethomepage.dev/latest/widgets/services/proxmox/

[dwilliss]

I got it to work. The instructions on the above link, step 6 has you set a Group Permission on the group created earlier to give it the PVEAuditor role. That didn't work for me. But what did was skipping that step and go ahead and create the API Token. Then go back and give the PVEAuditor role to the token itself.

Maybe this is something that is broken in the version of PVE I'm currently on (7.4-3) because if the token is owned by the user and the user is in the group and the group has the role and everything is set to propagate, it should just work as documented.

[jtfinley72]

Also having issues with Proxmox 8.1.3 (API Error information). Actually hard-coded credentials in and still errors. Tested via Postman so unsure what the issue is.

api@pve and different iterations of API key, same error. Restarted docker, etc. no joy.

[jtfinley72]

Silly me; reviewed the homepage site and updated info

widget:
type: proxmox
url: https://proxmox.host.or.ip:8006
username: api_token_id
password: api_token_secret
node: pve-1 # optional

[AdamXweb]

Nice guide!
+1 on the interest getting a .env working with the services!

[atw527]

I had the same problem. The .env file needs to be referenced in the docker-compose.yml file in order for this to work. Look at line 9 of Tim's docker-compose.yml example. Hope that helps!

[AdamXweb]

Do you mean the .env had variables for the docker compose file, or also secrets that homepage loaded for the services?

[atw527]

The .env file needs to be referenced in the docker-compose file. Look for this in the example: env_file: .env

[AdamXweb]

Ah right. I'm referring to the .env having secrets like API keys and logins for homepage services that would otherwise be in the services.yaml or other homepage configs

[mcbyte-it]

Ok, found the issue:
Homepage requires that the variable names in the .env file start with HOMEPAGE_VAR_, I thought that was optional and used shorted variable names.

[blackliner]

In case you use ArgoCD and Helm, I can recommend this chart: https://github.com/jameswynn/helm-charts/blob/main/charts/homepage/values.yaml
Make sure to use the latest image tag:

  image:
    repository: ghcr.io/gethomepage/homepage
    tag: "v0.8.4"

And instead of having a long list of all kubernetes host services (aka ingress), I decided to add the relevant annotations to allow Automatic Service Discovery. Example for OPNSense:

apiVersion: networking.k8s.io/v1
kind: Ingress
metadata:
  name: opnsense-ingress
  annotations:
    gethomepage.dev/enabled: "true"
    gethomepage.dev/name: "OPNSense"
    gethomepage.dev/group: "Automatic service discovery"
    gethomepage.dev/icon: "opnsense.png"
    gethomepage.dev/widget.type: "opnsense"
    gethomepage.dev/widget.url: "https://your.url.com"
    gethomepage.dev/widget.username: <KEY>
    gethomepage.dev/widget.password: <SECRET>
spec:
  ...

[spasmonaut]

I feel like something obvious is escaping me but after I've created the container, having already made a config directory within the directory where my compose file exists, and used ./config in the compose yaml for config volume. My config directory is always empty after starting the container, unlike https://youtu.be/mC3tjysJ01E?si=pgh5WcVzOMT1G-r1&t=386

[spasmonaut]

Sorted out my issue...apparently I was not waiting long enough for the creation of the container and its related components to take flight. In case anyone else runs into this...

[CH-MrMaus]

Yeah, Mine is Running ina LXC, on its own, with but one cpu core, activateed, took definatly more than 5 minutes, maybe 10, the first time..... Just so you do have a timeframe... If it is marked as unhealthy and the config folder is empty, just let it sit...

[mani76]

Hello,

Do you have example how to use providers to define secret key and then use that key in the services config?
I don't want use global docker environment variables.

[slapcat]

Thanks for the guide! Can you share how you wrote your uptime-kuma URL? I've been checking the docs and trying a bunch of different things, but nothing seems to get the data to populate in the widget. The tile just strobes on and off.

[dwilliss]

I just followed the instructions in the widget documentation.
The url for the widget is the same as the href. Don't add anything to it.
Then specify the slug. Mine looks like:

        widget:
            type: uptimekuma
            url: "http://10.0.0.49:3001"
            slug: "1"

Where the IP and port is where the docker container are exposed.

Actually, I guess my url is a bit different. In the href: I use https and a hostname that points to nginx proxy which then forwards to the actual docker service. If I do that in the widget url, it doesn't like it because homepage doesn't trust the CA that I used to sign my certificates and I haven't figured out how to add a trusted CA to a docker container. I doubt if that's your problem, but you never know.

[slapcat]

Thanks @dwilliss ! Last night I found my issue was the slug line. I was just using Tim's or the one of the docs because the docs weren't clear to me at the time.

As such you will need a status page setup with a group of monitored sites, which is where you get the slug (without the /status/ portion).

I thought it meant the URL shouldn't include the /status/ portion, not that the slug value is what comes after it.

[scottcopus]

I haven't used homelab yet. Does anyone know if it's possible to do change the homepage "root" url to be something NOT on the root... i.e. myserver.local/secret/ instead of myhome.local/ ? Anyone know what it reports in the HTTP headers as the "Server:" and can that be changed or removed? Thanks!

[brandmanNH]

Could you please give us an update on your homepage dahsboard? In your newest video (Gatus) it seems some things have changed since the post was written - I would be interessted how you added monitoring to your dashboard (the green dots in your services).
Thank you in advance!

[timothystewart6]

If you're looking for the gatus plugin for homepage, it's here https://gethomepage.dev/latest/widgets/services/gatus/ but if you're talking about the green uptime dots on each card in home page, you can do that by simply adding this to each section like so:

(This is a pihole example, they key parts are ping, and statusStyle)

    - DNS:
        - Pi-Hole1:
            icon: pi-hole.svg
            href: "{{HOMEPAGE_VAR_PIHOLE_URL_1}}"
            description: quasar
            ping: "{{HOMEPAGE_VAR_PIHOLE_HOST_1}}"
            statusStyle: "dot"
            widget:
                type: pihole
                url: "{{HOMEPAGE_VAR_PIHOLE_API_URL_1}}"
                key: "{{HOMEPAGE_VAR_PIHOLE_API_KEY_1}}"

[ryanspags]

I set up homepage with the LXC container from the Proxmox VE helper-scripts everything is working great except how to reference the .env file. Thanks for all the great videos!

[ryanspags]

never mind found my mistake. Needed to single quote ip addresses to avoid the api error.

[tejasthakur123]

Where to find .env file in LXC container from the Proxmox VE helper-scripts