Nothing to do for Prometheus #1576
Comments
|
@mgonzaga1990 thanks for reporting the issue. |
|
@balchua thanks, meanwhile, i tried accessing influxdb from however it is giving me. but i can access grafana |
|
Fyi the influxdb and heapster were removed from MicroK8s 1.19 channel. The kube dashboard comes with the scraper and dashboard alone. Coming back to your issue, what about port forwarding? |
|
@balchua scraper? sounds like prometheus to me? does that mean it has built in prometheus? how about grafana? will still be there? |
|
Sorry i may have confused you. The dashboard scraper is not prometheus. It is part of the dashboard pods to get some metrics. It doesn't come with prometheus nor grafana. The monitoring stack will still be the prometheus addon. Which is based on kube-prometheus. |
|
i see. but let me verify my understanding.. so will there be prometheus on 1.19? how about grafana will still be there?. Sorry I'm still exploring microk8s and planning to stick with it |
|
@balchua if you could tell me what to do i could help you test it in arm64 |
|
@mgonzaga1990 there is a file named - name: "prometheus"
description: "Prometheus operator for monitoring and logging"
version: "v0.6.0"
check_status: "pod/prometheus-k8s-0"
supported_architectures:
- amd64But this file is read only. I don't know how to mount it to a writable filesystem. |
|
I think that is different in my end.
cant see microk8s-resources |
|
@balchua i think as of now i may have to manually deploy prometheus in my cluster.. what image would you recommend for arm64? |
|
@mgonzaga1990 this is how we package the kube-prometheus https://github.com/ubuntu/microk8s/blob/master/microk8s-resources/actions/enable.prometheus.sh |
|
@balchua thanks for this.. i'm planning to upgrade now.. btw. what is the difference between classic and stable? what would you recommend for raspberry pi? |
|
@mgonzaga1990 you can find here more information about the channels. |
|
@balchua i think i have to stick with 1.18 as of now.. i tried using the 1.19 but at some point my raspberry pi freezes when running pipeline (drone and gitea) which i don't experience with 1.18 |
|
If it is not too much, will it be possible to get the inspect tarball? Will really help to troubleshoot it. |
|
@balchua sure sure.. i'll geenrate it and let you know once available |
|
@mgonzaga1990 perhaps opening a new issue with the 1.19 freezing up your rpis will be a good way to track that. Wdyt? |
|
yup sure... ill set up my cluster back to 1.19
…On Tue, Sep 22, 2020, 6:58 PM balchua ***@***.***> wrote:
@mgonzaga1990 <https://github.com/mgonzaga1990> perhaps opening a new
issue with the 1.19 freezing up your rpis will be a good way to track that.
Wdyt?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#1576 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AFBKRM7IETYGC67FPY2JNF3SHB7PDANCNFSM4RPGQ7BQ>
.
|
|
I have the same problem on the 1.19 channel. I'm on a pi 4b and ubuntu 20.04.1. I think the arm64 issue may be a red herring. I am new to microk8s and snap, but looking through I ran Here's the output: In the pods that were created there are still several problems. Specifically the logs of several crashing pods mention |
|
Following up from the link to prometheus-operator/kube-prometheus#545 I removed microk8s 1.19 and intalled 1.18 (I hadn't noticed that kube-prometheus support for 1.19 was still at HEAD and not 0.6). I then followed the same procedure as I outlined in my previous post and it failed with I ran through the commands in The next step failed with the same path problem. But I changed it to The pods that start up still have the same FWIW, cloning the kube-prometheus repo and running through the Quickstart resulted in prometheus-operator coming up successfully, followed by node-exporter, grafana, prometheus-adapter, alert-manager-main-{0,1,2}, prometheus-k8s-{0,1}. However the kube-state-metrics container still fails with |
|
As a workaround on 1.19 you can do as suggested in kubernetes/kube-state-metrics#1190 (comment) and modify https://github.com/prometheus-operator/kube-prometheus/blob/980e95de011319b88a3b9c0787a81dcdf338a898/manifests/kube-state-metrics-deployment.yaml#L26 from to and everything comes up as expected. |
|
I can confirm that the prometheus add-on installed successfully in microk8s 1.20 on my raspberry pi cluster running Ubuntu 20.04.1 (like the others in this thread I had problems with 1.19). Given that, I think this issue can now be closed. |
|
Thanks @seangustafson, this addon also works on mixed clusters. I.e a mix of arm64 and amd64 nodes. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I tried enabling prometheus on my microk8s cluster on raspberry pi however it is giving me this below result.
It seems prometheus is not part of the default addons
I can access grafana
There is influxdb however, we like to use prometheus as we have experience on that.
I'm following this example but since there is no prometheus it looks like i cant
https://ubuntu.com/blog/monitoring-at-the-edge-with-microk8s
lastly this is my microk8s version
--classic --channel=1.18/stableThe text was updated successfully, but these errors were encountered: