Dashboard: kubernetes-metrics-scraper on ARM on 2.0beta1: Unable to initialize database tables: Binary was compiled with 'CGO_ENABLED=0

[Rick-Jongbloed]

Environment

• Kubernetes Cluster with Raspberry PI 3 / 4
• ARM architecture
• Installation method: alternative deployment yaml for 2.0 beta 1 (kubectl apply -f .....)
• Kubernetes version: K3S (https://k3s.io/)

Server Version (outpunt from: kubectl version): version.Info{Major:"1", Minor:"14", GitVersion:"v1.14.3-k3s.1", GitCommit:"8343999292c55c807be4406fcaa9f047e8751ffd", GitTreeState:"clean", BuildDate:"2019-06-12T04:56+00:00Z", GoVersion:"go1.12.1", Compiler:"gc", Platform:"linux/arm"}

• Dashboard version: 2.0beta1
• Operating system: Debian Stretch 9.9
• Node.js version ('node --version' output): Node.js is not installed on the nodes
• Go version ('go version' output): go1.12.1

Steps to reproduce

1. apply alternative yaml file to the ARM kubernetes cluster (https://raw.githubusercontent.com/kubernetes/dashboard/v2.0.0-beta1/aio/deploy/alternative.yaml)
2. Log on to the dashboard after the dashboard is loaded.
3. The pod for kubernetes-metrics scraper is in error state.
4. The log file shows:

2019-07-07T19:52:32.38087562+01:00 {"level":"info","msg":"Kubernetes host: https://10.43.0.1:443","time":"2019-07-07T18:52:32Z"} 2019-07-07T19:52:32.382007209+01:00 {"level":"fatal","msg":"Unable to initialize database tables: Binary was compiled with 'CGO_ENABLED=0', go-sqlite3 requires cgo to work. This is a stub","time":"2019-07-07T18:52:32Z"}

Observed result

See above.

Expected result

The kubernetes-metrics scraper is in running state.

Comments

I'm not really sure why the error occurs. If you need more infomation. please let me know.

[michael-robbins]

Getting this as well on my cluster of Raspberry Pi 3's

Environment

• Kubernetes Cluster with Raspberry PI 3
• ARM architecture
• Installation method: alternative deployment yaml for 2.0 beta 1 (kubectl apply -f .....)
• Kubernetes version: k8s 1.15.0

Server Version (output from: kubectl version):

Client Version: version.Info{Major:"1", Minor:"15", GitVersion:"v1.15.0", GitCommit:"e8462b5b5dc2584fdcd18e6bcfe9f1e4d970a529", GitTreeState:"clean", BuildDate:"2019-06-19T16:40:16Z", GoVersion:"go1.12.5", Compiler:"gc", Platform:"linux/arm"}
Server Version: version.Info{Major:"1", Minor:"15", GitVersion:"v1.15.0", GitCommit:"e8462b5b5dc2584fdcd18e6bcfe9f1e4d970a529", GitTreeState:"clean", BuildDate:"2019-06-19T16:32:14Z", GoVersion:"go1.12.5", Compiler:"gc", Platform:"linux/arm"}

Dashboard version: 2.0beta1
Operating system:

pi@node00:~ $ lsb_release -a
No LSB modules are available.
Distributor ID: Raspbian
Description:    Raspbian GNU/Linux 9.9 (stretch)
Release:    9.9
Codename:   stretch

Node.js version ('node --version' output): N/A
Go version ('go version' output): N/A

[floreks]

@jeefy can we rebuild scraper for arm arch? Is it related to some dep or our build process?

[jeefy]

Huh, interesting. Our build currently does multiarch and is doing builds for arm and arm64 but this might be an unforeseen failure in the sqlite library we use on arm. I've got a Pi3 at home I can play with, but it might not be till this weekend.

/assign

Quick aside, docker manifest inspect kubernetesdashboarddev/dashboard-metrics-sidecar outputs all the different targets the container image will work on (if you have docker cli experimental enabled). :)

[MattJeanes]

Also getting this

Same as the others, Raspberry Pi cluster ARM32

[mmb-davidsmith]

I'm also seeing the same on an arm cluster:

{"level":"info","msg":"Kubernetes host: https://10.96.0.1:443","time":"2019-07-09T20:30:39Z"}
{"level":"fatal","msg":"Unable to initialize database tables: Binary was compiled with 'CGO_ENABLED=0', go-sqlite3 requires cgo to work. This is a stub","time":"2019-07-09T20:30:39Z"}

I installed as per the instructions at https://kubernetes.io/docs/tasks/access-application-cluster/web-ui-dashboard/, using the recommended install

[jeefy]

Have a PR in flight to fix this. :) https://github.com/kubernetes-sigs/dashboard-metrics-scraper/pull/15

[Rick-Jongbloed]

I believe you fixed it.

image: kubernetesdashboarddev/dashboard-metrics-sidecar

Logging is now
{"level":"info","msg":"Kubernetes host: https://10.43.0.1:443","time":"2019-07-16T11:20:16Z"}

And stats are shown on the dashboard. Thanks!

[lassebm]

@jeefy Can I ask you to fix this for arm64 as well? Thanks!

[charles-d-burton]

This is still happening on on arm64.

[jeefy]

Reopening to track arm64 (hopefully won't stay reopened for long)

[zimme]

https://github.com/kubernetes-sigs/dashboard-metrics-scraper/releases/tag/v1.0.2

It seems as though they finally released a new version of the metric scraper with the arm64 fix :tada:
I guess we just need a new release which is using version 1.0.2 of the metric scraper to test this out.

[floreks]

This isn't ready yet. I started preparing Travis pipeline to automatically release scraper when GitHub release is created. Still, we have to fix the configuration. Once release notes will be there, it will be ready to use. You won't have to wait for us to release Dashboard. You can simply bump scraper to v1.0.2 manually once it will be pushed to docker.

[floreks]

@Rick-Jongbloed @zimme Scraper v1.0.2 should be now working with ARM64. https://github.com/kubernetes-sigs/dashboard-metrics-scraper/releases/tag/v1.0.2

Reopen if there are still some issues.

/close

[zimme]

kubernetes-dashboard up and running on arm + arm64 dev board cluster running k3os + k3s :tada:

[jonstelly]

I'm getting this error. I just deployed beta 8 on k3s arm64 with kubectl apply -f https://raw.githubusercontent.com/kubernetes/dashboard/v2.0.0-beta8/aio/deploy/recommended.yaml but recommended.yaml still uses kubernetesui/metrics-scraper:v1.0.1.

I'm not sure of the process, but does the recommended.yaml need a manual update to v1.0.2?

[zimme]

@jonstelly I had to manually update recommended.yaml to use 1.0.2.

[jonstelly]

Yeah, I forgot to mention I did the same and everything's working. I just wanted to bump this to make sure it made it into the next beta.

[kamikaze]

having same issue on rpi4

[dawidslowik]

I'm also having this issue on rpi4 with ubuntu 19.10 64bit. Downgrading kubernetesui/metrics-scraper from 1.0.4 to 1.0.2 helps