devspace-cloud / devspace-cloud Goto Github PK

When I run the command in README to add a new provider, devspace ask to add first.

# devspace use provider devspace.domain.net.br -n devspace-cloud
[fatal]  Error provider devspace.domain.net.br does not exist! Did you run `devspace add provider devspace.domain.net.br` first?

But when I try to add with the access key:

[info]   Unable to open web browser for login page.
Please follow these instructions for manually loggin in:

  1. Open this URL in a browser: https://devspace.domain.net.br/settings/access-keys
  2. After logging in, click the 'Create Key' button
  3. Enter a key name (e.g. my-key) and click 'Create Access Key'
  4. Copy the generated key from the input field

? 5. Enter the access key here: ********************************

[fatal]  log into provider: Login: login: get token: token save: Couldn't find provider devspace.domain.net.br

I'm using Azure AKS and I can acess with kubectl. I followed the steps, but always get the same error. Please help.

Regards

Currently, when removing a cluster via UI which is not existing anymore, the user gets an error message that the API server host cannot be resolved/reached. We should offer a "force remove" option for this case because it is otherwise not possible to remove clusters which do not exist anymore.

This issue may also exist for Spaces.

Hi,

It seems we can't add external providers anymore since 4.4.0.

I tried doing : devspace add provider devspace.xxx.com (my custom devspace provider, that is accessible).

The command returns : Cloud provider not found! Did you run 'devspace add provider [url]'.

It's only the case for new members of the team we have here. Note: devspace.xxx.com is accessible.

Thanks!

Following registry of 'devspace-cloud' used was not found in hub.docker.com
devspacecloud/auth:0.0.6

devspacecloud/hasura:0.0.4

devspacecloud/hasura-init:0.0.5

There should be an option to disable the nginx ingress controller that gets deployed with devspace-cloud.

Another option is to remove it from the chart entirely, I would think it is pretty common for a cluster to already have an ingress controller set up, it seems a bit out of scope for this as well.

Why not just let whatever ingress controller is set up as the default handle any created ingresses?

extensions/v1beta1/Ingress is working but not networking.k8s.io/v1beta1/Ingress which is stated in several examples within the k8s documentation => Error message:

rpc error: code = Unknown desc = validation failed: unable to recognize "": no matches for kind "Ingress" in version "networking.k8s.io/v1beta1"

In addition to my issue for DevSpace UI, I was wondering whether a search bar could be added for the DevSpace Cloud space view too.

I think for most tabs in the space view, it would be very helpful to have search bar, possible right under the navigation tab, to search for container names / event texts / etc. Currently, I'm using the browser's 'find in page' functionality, but with open logs and a container name that is written in the logs, this is quite cumbersome to do :].

Would love to hear what you think :].

What happened?

UI shows error get account: record not found under certain circumstances (= session data stored in local storage from old instance on same domain).

What did you expect to happen instead?

Redirect user to login if session is not valid or user is not found or any other auth error occurs.

How can we reproduce the bug? (as minimally and precisely as possible)

1. Install DevSpace Cloud via devspace run deploy-devspace-cloud-no-tls
2. Sign up and login.
3. Uninstall DevSpace Cloud via devspace purge -s
4. If local cluster, also purge local storage data via devspace run purge-local-storage
5. Install DevSpace Cloud again via devspace run deploy-devspace-cloud-no-tls (do not reset vars or use the same domain name)
6. Wait until everything is running and open: http://${DOMAIN}/spaces

Outline:

• install CLI
• connect clusters
• configure important limits
• create and manage spaces
• create invite links and manage users

Hello,
I am encountering a problem when first installing DevSpace On-Premise Edition via Helm.

Error:
"Error: template: devspace-cloud/templates/auth/deployment.yaml:49:22: executing "devspace-cloud/templates/auth/deployment.yaml" at <.Values.auth.liveness>: nil pointer evaluating interface {}.liveness"

Steps to reproduce:

• Launch a one-node-cluster with RKE on Ubuntu18.04
• Install DevSpace On-Premise with the provided helm command in your documentation.

There is a open issue at helm regarding this, but do you have a workaround for this problem?

Thanks very much!

Currently, https connections are allowed even when DevSpace Cloud is deployed as http-only and this leads to issues with the session storage. If the users logs in via https and adds DevSpace Cloud as provider via devspace add provider --host=http://... using http, the CLI will open the login page because the user is not authenticated because the authentication is limited to the https version.

1. If HTTPS is configured, requests should always be forwarded to HTTP (but NON-permanent redirect, so status code 307 instead of 301 because otherwise switching the protocol will never be possible due to infinite browser caching of 301 redirects).
2. If HTTP is configured, requests to HTTPS should not be accepted or forwarded as 307 to HTTP as well. However, it is important that the ingress controller itself is accepting HTTPS requests, e.g. for space domains etc.

We should publish the devspace cloud chart to a chart repository and add a README section how to deploy via plain helm.

It would be nice if devspace-cloud's authentication worked with Active Directory at some point down the road.

Hi,

I created a space, that created a namespace in Kubernetes.
A few days later I've deleted that namespace with kubectl delete ns ....

Now the space is still existing and I can't devspace use space .... because the serviceaccount doesn't exist anymore.
But I also can't delete the space from devspace cloud. (Internal server error)

Thanks!

Example error:

rpc error: code = Unknown desc = failed to create resource: admission webhook "limits.devspace.cloud" denied the request: Ingress my-component annotation helm.sh/chart is not allowed

running:
devspace run deploy-devspace-cloud-no-tls

Currently sets --var DOMAIN="" and I'm not sure why, but it's causing the ingress to have no host value.

I'll create a PR to fix this momentarily.

[fatal] Error deploying devspace-cloud: Unable to deploy helm chart: helm install: rpc error: code = Unknown desc = release devspace-cloud failed: Deployment.apps "devspace-ui" is invalid: [spec.selector: Required value, spec.template.metadata.labels: Invalid value: map[string]string{"app":"devspace-ui"}: `selector` does not match template `labels`]