# Grafana Cloud Node Exporter metrics gone after helm upgrade (kubernetes / k3s)

**URL:** <https://community.grafana.com/t/grafana-cloud-node-exporter-metrics-gone-after-helm-upgrade-kubernetes-k3s/114563>\
**Category:** Grafana Cloud\
**Tags:** grafana-cloud, kubernetes\
**Created:** [February 15, 2024, 11:05am UTC](https://community.grafana.com/t/grafana-cloud-node-exporter-metrics-gone-after-helm-upgrade-kubernetes-k3s/114563 "2024-02-15T11:05:38Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![olafkocanda](https://sea2.discourse-cdn.com/grafana/user_avatar/community.grafana.com/olafkocanda/32/21537_2.png) [@olafkocanda](https://community.grafana.com/u/olafkocanda)\
**Post date:** [February 15, 2024, 11:05am UTC](https://community.grafana.com/t/grafana-cloud-node-exporter-metrics-gone-after-helm-upgrade-kubernetes-k3s/114563/1 "2024-02-15T11:05:38Z")

</div>

Hi there, I am kind of new to this so please be patient 🙂

I have a k3s cluster on my rpi and just recently upgraded the grafana monitoring metrics via helm

`helm upgrade grafana-k8s-monitoring grafana/k8s-monitoring -n "monitoring"`

Interestingly, now the `node_uname_info` call which is needed for the variables in grafana seems to stop being sent to Grafana cloud:

 ![ptfe3ot49qic1](https://us1.discourse-cdn.com/grafana/original/3X/3/b/3ba6840d822013be9cc5eadc5c79ab2f7c9af2ba.webp)

When grepping them locally on my rpi it still exists, even though labels like “instance” are not there (i guess that is a kubernetes config setting).

```auto
curl http://localhost:9100/metrics | grep uname
% Total % Received % Xferd Average Speed Time Time Time Current` ` Dload Upload Total Spent Left Speed 100 133k 0 133k 0 0 814k 0 --:--:-- --:--:-- --:--:-- 814k
node_scrape_collector_duration_seconds{collector="uname"} 6.174e-05
node_scrape_collector_success{collector="uname"} 1` `# HELP node_uname_info Labeled system information as provided by the uname system call.` `# TYPE node_uname_info gauge` 
node_uname_info{domainname="(none)",machine="aarch64",nodename="raspberrypi",release="5.10.103-v8+",sysname="Linux",version="#1529 SMP PREEMPT Tue Mar 8 12:26:46 GMT 2022"} 1

```

At first I thought the connection to grafana cloud is misconfigured after the upgrade but the metrics status via `node_exporter_build_info` is still there:

 ![iwmbbpg8cqic1](https://us1.discourse-cdn.com/grafana/original/3X/2/6/26b753061637ee1b03d517e415a2655bee2d9479.webp)

Am I missing something essential here?

Thanks for any help!

---

<div class="post-metadata">

**Author:** ![olafkocanda](https://sea2.discourse-cdn.com/grafana/user_avatar/community.grafana.com/olafkocanda/32/21537_2.png) [@olafkocanda](https://community.grafana.com/u/olafkocanda)\
**Post date:** [February 15, 2024, 2:41pm UTC](https://community.grafana.com/t/grafana-cloud-node-exporter-metrics-gone-after-helm-upgrade-kubernetes-k3s/114563/2 "2024-02-15T14:41:31Z")

</div>

Update: In the meantime I deleted all services, configmaps and running statefulsets, daemonsets, deployments, deleted unnecessary helm repos and started again the official way:

 ![grafik](https://us1.discourse-cdn.com/grafana/original/3X/6/0/601e7e1dc2ed6380dda4d4625972087d221016ba.png)

Now while it seems to have some of the metrics imported (see node fleet overview dashboard):

 ![grafik](https://us1.discourse-cdn.com/grafana/original/3X/e/f/ef559672ab0d7cff1a6faf205fa44bf64529e403.png)

However, there is still nothing for `node_uname_info`…

---

<div class="post-metadata">

**Author:** ![petewall](https://sea2.discourse-cdn.com/grafana/user_avatar/community.grafana.com/petewall/32/32469_2.png) [@petewall](https://community.grafana.com/u/petewall)\
**Post date:** [February 15, 2024, 3:25pm UTC](https://community.grafana.com/t/grafana-cloud-node-exporter-metrics-gone-after-helm-upgrade-kubernetes-k3s/114563/3 "2024-02-15T15:25:47Z")

</div>

Hello! Thanks for your post.

I am one of the authors of the k8s-monitoring Helm chart.

The chart has a default set of metrics that it allows to be saved and sent for storage. By default, it only keeps a subset of the Node Exporter metrics ([k8s-monitoring-helm/charts/k8s-monitoring/default\_allow\_lists/node\_exporter.yaml at main · grafana/k8s-monitoring-helm · GitHub](https://github.com/grafana/k8s-monitoring-helm/blob/main/charts/k8s-monitoring/default_allow_lists/node_exporter.yaml)). However, there is an easy switch to include a larger set of metrics to work with the Node Exporter integration and dashboards. Here is the larger set: [k8s-monitoring-helm/charts/k8s-monitoring/default\_allow\_lists/node\_exporter\_integration.yaml at main · grafana/k8s-monitoring-helm · GitHub](https://github.com/grafana/k8s-monitoring-helm/blob/main/charts/k8s-monitoring/default_allow_lists/node_exporter_integration.yaml). Note that the node\_uname\_info metric is included in that list.

To enable that set, add this to your values file:

```auto
metrics:
  node-exporter:
    metricsTuning:
      useIntegrationAllowList: true

```

That should have the Grafana Agent include the larger set of metrics and then it should show up inside Grafana!

---

<div class="post-metadata">

**Author:** ![olafkocanda](https://sea2.discourse-cdn.com/grafana/user_avatar/community.grafana.com/olafkocanda/32/21537_2.png) [@olafkocanda](https://community.grafana.com/u/olafkocanda)\
**Post date:** [February 15, 2024, 5:39pm UTC](https://community.grafana.com/t/grafana-cloud-node-exporter-metrics-gone-after-helm-upgrade-kubernetes-k3s/114563/4 "2024-02-15T17:39:14Z")

</div>

Thanks so much!  
This solved the issue:

 ![grafik](https://us1.discourse-cdn.com/grafana/original/3X/e/c/ecffcc3e5f06ed1e3b2597e6265cf5a7e3014f13.png)

For anyone else stumbling across here. To update the values.yaml for an existing file I used:

```auto
helm upgrade --reuse-values -f git/grafana/helm/values.yaml grafana-k8s-monitoring grafana/k8s-monitoring --version 0.10.2 --namespace monitoring

```

This let’s you keep the version. The version can be found via `helm search repo grafana/k8s-monitoring`