kubero-dev / kubero Goto Github PK

Since it requires connecting the repository, it is actually not possible to deploy a public, not owned repository.

This should be made possible.

We need a overview how Kubero is configured.

The docs are missing
https://code.onedev.io/help/api

https://code.onedev.io/onedev/manual/~files/main/pages/restful-api.md

Please point me, where to the API Docs. I was not able to find them, since I am not very familiar with oneDev.

Investigating

Version < v1.3.0

[1] listRepos: gitea
[1] Response {
[1]   size: 0,
[1]   timeout: 0,
[1]   data: null,
[1]   error: FetchError: invalid json response body at http://localhost:3000/api/v1undefined reason: Unexpected token < in JSON at position 0
[1]       at /kubero/node_modules/node-fetch/lib/index.js:273:32
[1]       at processTicksAndRejections (node:internal/process/task_queues:96:5)
[1]       at async/kubero-dev/kubero/node_modules/gitea-js/dist/index.js:145:44 {
[1]     type: 'invalid-json'
[1]   },
[1]   [Symbol(Body internals)]: {
[1]     body: PassThrough {
[1]       _readableState: [ReadableState],
[1]       _events: [Object: null prototype],
[1]       _eventsCount: 4,
[1]       _maxListeners: undefined,
[1]       _writableState: [WritableState],
[1]       allowHalfOpen: true,
[1]       [Symbol(kCapture)]: false,
[1]       [Symbol(kCallback)]: null
[1]     },
[1]     disturbed: true,
[1]     error: null
[1]   },
[1]   [Symbol(Response internals)]: {
[1]     url: 'http://localhost:3000/api/v1undefined',
[1]     status: 404,
[1]     statusText: 'Not Found',
[1]     headers: Headers { [Symbol(map)]: [Object: null prototype] },
[1]     counter: 0
[1]   }
[1] }

create a job to run trivy on the local files and images.

How to check :

kubectl get pods --all-namespaces | grep kubero-operator-controller-manage

Results in following error:

kubero-operator-system   kubero-operator-controller-manager-XXXXXXXXX   0/2     ErrImagePull        0          3m8s

Is the header for the website supposed to look like this?

https://i.imgur.com/VmHMieM.png

Same result in Firefox and Brave (so I assume Chrome is the same). This seems to be happening only on wide views, about 992px and up.

The links on the addon name are useless.

There should be also a docs link directly to kubero logs.

Would love to have Hugo as a buildpack for Kubero.

you already have the buildpack, just need to add it to the API for pulldown :)

If you are using external volumes you may see pods stuck in the init fetch container after an automatic git push deployment or restart.

This is because the extra volume can not be mounted twice. The PVC (Persistant Volume Claim) is created as RWO ( ReadWriteOnce )

Not sure how to fix this yet. Changing the deployment strategy ( which would not be so cool, since it may lead to service disruptions during deployment) or changing extra volumes to RWM (ReadWriteMany)

Currently when setting up different branches for each environment they always use main.

Tracking issue for:

• https://github.com/kubero-dev/kubero/security/code-scanning/26

can we get a table or list of minimum required permissions when generating tokens for each git provider

Kubero version: 1.4.0
Quickfix :
Edit with kubectl the value in kuberoapps.application.kubero.dev deploymentstrategy: git to deploymentstrategy: docker

Command :

kubectl edit kuberoapps.application.kubero.dev

When setting up an app in the UI, I turn web pods to 0 and worker pods to 1 but web pods still return 1 when the app generates the Kuberoapp .

Although as far as I understand from docs they are not, but it will be great if we could implement deploy as highly available system, which basically tries to create 3 copies of application with pod affinity to deploy on different nodes ensuring high availability

I've seen some issues deploying PHP Applications.

The current running image is not permissive enough.

I decided to go with Jest. Other Ideas or suggestions are welcome.

Currently I am setting up a nodejs app and not listening on any ports. My deployments don't go into a ready state and restart because the liveness and readiness timeouts. Can liveness and readiness probes be configured or turned off ?

Will fix this today

09:27:42 - Starting compilation in watch mode...
[0]
[1] socket.io started
[1] load kubectl config from file
[1] Gitlab API: undefined
[1] Gitlab token: glpat-.......
[1]  BITBUCKET CLOUD API LATEST UPDATES:  https://developer.atlassian.com/cloud/bitbucket
[1] /Users/carafagi/workdir/kubero-dev/kubero/node_modules/bitbucket/lib/index.umd.js:15
var r}function W(e){var r,t;return!1!==U(e)&&("function"==typeof(r=e.constructor)&&(!1!==U(t=r.prototype)&&!1!==t.hasOwnProperty("isPrototypeOf")))}function F(){var e=arguments.length>0&&void 0!==argum

We need a check box for every app to enable TLS support with cert-manager headers.

Operator and CLI are prepared already.

Version 0.0.1 seems to have an issue connecting the repo caused by the failing webhook.

Looking into that asap.

This is a minor bug.

I had to take down the Demo (demo.kubero.dev) due to some issues with persistent volumes PV.

Affects only Version v1.5.0

The Operator does not check vor installed resource version :

"failed to get candidate release: unable to recognize \"\": no matches for kind \"HorizontalPodAutoscaler\" in version \"autoscaling/v2beta1\"

In some cases it is autoscaling/v2

to check which resource you have use kubectl api-resources

{"level":"error","ts":1671789873.250903,"logger":"helm.controller","msg":"Failed to sync release","namespace":"customcommand-stage","name":"tls","apiVersion":"application.kubero.dev/v1alpha1","kind":"KuberoApp","release":"tls","error":"failed to get candidate release: unable to recognize \"\": no matches for kind \"HorizontalPodAutoscaler\" in version \"autoscaling/v2beta1\"","stacktrace":"sigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Reconcile\n\t/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:114\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).reconcileHandler\n\t/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:311\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).processNextWorkItem\n\t/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:266\nsigs.k8s.io/controller-runtime/pkg/internal/controller.(*Controller).Start.func2.2\n\t/go/pkg/mod/sigs.k8s.io/[email protected]/pkg/internal/controller/controller.go:227"}

cant read fetch

Heroku still does not provide a persistent storage for their pods.

But many apps require it to run. So why don't add it?

currently, the UI for GH repo's shows a static set of repos, it should show the correct ones.

For any repo, it shows the following branches:

dumm-pr2
dumm-pr
main
noProc

It will default to main

Kubero Version : v1.3.1
Kubero Operator Version : v0.0.74

restart app: test
restart app: test
ERROR: deployments.apps "test-kuberoapp-web" is forbidden: User "system:serviceaccount:kubero:kubero-sample" cannot patch resource "deployments" in API group "apps" in the namespace "test-production"
ERROR: HttpError: HTTP request failed
ERROR: deployments.apps "test-kuberoapp-worker" is forbidden: User "system:serviceaccount:kubero:kubero-sample" cannot patch resource "deployments" in API group "apps" in the namespace "test-production"
ERROR: HttpError: HTTP request failed

The branches are currently a text box prefilled with "main". This should be loaded dynamically when the "new app" page is loaded.

Upon a new app creation

Web requests still POST to https://kubero.local.domain/api/apps

Making an PUT request on the frontend to https://kubero.local.domain/api/pipelines/pipeline1/production/abc works correctly

Relates to #48

Docs schould be moved from wiki to /docs to make them editable

We need a dark mode. The default is to choose by client hour. If selected it should be stored in a cookie.

Apps require a wildcard path "/*" in ingress configuration but got only "/"

I am testing some solutions for that

For someone who is not familiar with the project, this deployment strategy toggle might prove to be a bit confusing.

At first I thought that this means that docker deployments are turned off when that is not the case.

Maybe a radio button or something else would make more sense here to distinguish between docker/git deployments?

I'd like to add some basic stats. This will require adding the metrics server :
https://github.com/kubernetes-sigs/metrics-server

But this was required by HPA anyway.