Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Portal and argo-cd login both (intermittently) lead to blank pages. #12

Open
davedavemckay opened this issue Dec 6, 2024 · 4 comments
Labels
bug Something isn't working

Comments

@davedavemckay
Copy link
Collaborator

  1. On clicking the "Portal" button on the RSP homepage, the user is sometimes taken to a blank page.
  • The user must navigate back, then forward again to see the Portal.
  1. On login to argo-cd, the user is sometimes taken to a blank page.
  • The user must close their browser tab, then open a new one and navigate to argo-cd for it to load correctly.

It has been suggested this is a LoadBalancer issue, which may correlate with it being intermittent.

@davedavemckay davedavemckay added the bug Something isn't working label Dec 6, 2024
@stvoutsin
Copy link
Collaborator

To see if the load-balancer tempfix works you can:

  • Navigate to Network / Load Balancers in the Openstack Dashboard
  • Find and click on the lb that is used for the cluster we want to fix (e.g. ..something-prod..)
  • For each of the pools in the lb, delete the members that aren't running the nginx controller pod. Normally only 1 ingress controller pod is run, unless you run it as a daemonset or have multiple replicas, so you'd only leave the master node & the 1 node running the ingress pod.

I think we had also experimented with Amphora and a health-checker, which handled the above, which I recall working though I'm not sure about it.
(see: https://github.com/lsst-uk/rsp-uk-docs/wiki/RSP-Deployment-instructions-on-Openstack-with-Magnum#setup-amphora-loadbalancer-healthchecker)

In any case if this is the cause the load balancer configuration probably needs to be revisited.

@davedavemckay
Copy link
Collaborator Author

davedavemckay commented Dec 6, 2024

Thanks @stvoutsin
I only see two pools:

pool_0_kube_service_rsp-prod-mint-7j4bka3i6afe_ingress-nginx_ingress-nginx-controller
pool_1_kube_service_rsp-prod-mint-7j4bka3i6afe_ingress-nginx_ingress-nginx-controller

@gpfrancis has mentioned issues with Amphora. We're currently using OVN.

@stvoutsin
Copy link
Collaborator

Actually I think we recently found a better workaround which was to deploy the ingress nginx as a Daemonset, which Amanda deployed on test, though I don't know if that fixed made it to main/prod:

See: lsst-sqre/phalanx@main...lsst-uk:phalanx:rsptest

Assuming this is deployed I would expect the LB issue to be resolved hopefully

@stvoutsin
Copy link
Collaborator

I tried logging into argo-cd and get an incorrect redirect to:
https://rsp-test.lsst.ac.uk/argo-cd/argo-cd/applications
instead of:
https://rsp-test.lsst.ac.uk/argo-cd/applications
So wherever the redirect is defined is broken somehow. Could you try syncing argo-cd to the latest version and trying again?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working
Projects
None yet
Development

No branches or pull requests

2 participants