Hi,

I'm a happy user of your addon, but at the moment, I'm a bit reluctant to update to the latest version, because it's named 'dev' (also see the code here). Wouldn't that cause issues in a subsequent update?

Regards,
Bert

This issue lists Renovate updates and detected dependencies. Read the Dependency Dashboard docs to learn more.

Open

These updates have all been created already. Click a checkbox below to force a retry/rebase of any.

• Update ghcr.io/hassio-addons/base/aarch64 Docker tag to v15.0.7
• Update ghcr.io/hassio-addons/base/amd64 Docker tag to v15.0.7
• Update ghcr.io/hassio-addons/base/armv7 Docker tag to v15.0.7
• Click on this checkbox to rebase all open PRs at once

Detected dependencies

• Check this box to trigger a request for Renovate to run again on this repository

Troubleshooting why the "Root FS Used" Gauge and "Total RootFS" Stat shows "N/A". I copied the query over to the Grafana "Explore" pane and was experimenting with the options there. I found that the metrics from the home assistant node_exporter integration does not have a "/" mountpoint. This is causing that gauge to read "N/A" and the query to return no data.

The relevant portion of the node_exporter gauge query is:
node_filesystem_avail_bytes{instance="$node",job="$job",mountpoint="/",fstype!="rootfs"}

If I change that query to be:
node_filesystem_avail_bytes{instance="$node",job="$job",mountpoint="/tmp",fstype!="rootfs"}

OR
node_filesystem_avail_bytes{instance="$node",job="$job",fstype!="rootfs"}

I get data back. Here is the job configuration in prometheus.yml

  - job_name: 'homeassistant'
    static_configs:
      - targets: ['homeassistant.home:9100']

I don't know if this integration isn't exporting the "/" mountpoint, or the integration doesn't have access to that mountpoint, or it's something else. I've attached the logs from the add on below along with a screen shot of the Grafana Explore query builder.

ha_node_exporter-startup.log.txt

Hi Logan, is this still working?
I've added your repo to an up to date version of HA but for some reason I cannot find your addon when I search for it?
Mike

Sorry folks, more of a question than an issue. But I'm trying to pinpoint what version of node_exporter this addon runs?

See this
renovatebot/renovate#18380

Issue with having multiple home assistant servers in multiple locations, the node name reports 0d869efa-prometheus-node-exporter for all node names.

Would you accept a PR adding new configuration options for CLI args?

The easy option would be to add a text field that directly contains CLI args:
CLI Args: --collector.textfile.directory="/root/config/prometheus_node_exporter/textfile_collector"
This is however not very user friendly. i could also add a new text field:
Textfile Directory: /root/config/prometheus_node_exporter/textfile_collector

But then we would have to add lots and lots of options for every possible argument.

Hi,

i try to run the addon on my ODroid-N2 running

Home Assistant Core 2022.5.4
Home Assistant Supervisor 2022.05.1
Home Assistant OS 7.6
Kernel version 5.10.108
Agent version 1.2.1

I did disable protection mode, however, starting the Addon i just end up with

s6-overlay-suexec: fatal: can only run as pid 1.

as sole entry in the log.

Anything i can do about this?

Thanks
Michael

Problems

1. Releases are not tagged. Everything is built from master when users click "install".
2. When users click "install", their HA instance needs to build the docker image in real-time. I would be better for me to pre-build images and publish them to GHCR.

Solutions

1. Start tagging releases. This doesn't do anything for users by itself, but will allow me to "roll back" to a specific "version" if I break something.
2. Per this comment and from this doc, I should publish images to GHCR. This would allow me to control what version of the add-on users receive and roll back if something breaks.

If you are an addon developer publishing addons that others use (and not just for personal use) I would strongly recommend publishing images for your addons to a container registry like ghcr.io or docker hub and then using the image option. This way existing versions of your addons will always keep working since the image is already built and available and not built on demand.