We now only fetch details for clusters that are currently in view. Improving the load on our backend. (#882)
The cluster detail view now redirects to the cluster list when the cluster you are viewing is no longer able to be found, for example when someone else deletes it. (#815)
We’ve replaced a warning mentioning that we do not support or monitor clusters below three worker nodes (which is not the case) with a recommendation to have at least three worker nodes. (#865)
We’re adding stricter linting and continue to improve the codebase consistency on various fronts.
Node Pool Support On installations that have node pools enabled (this is still a preview feature) Happa will show a new cluster creation form where you can create and edit node pools. The cluster detail screen is also updated, so that you can view and scale node pools after creating a cluster.
Cluster Deletion Date When deleting a cluster, our systems need time to fully remove all related resources. Sometimes deletion can be blocked due to an error or resource that we are unable to clean up. To help make this visible we’ve added the deletion date to the cluster list on the organization detail page. Clusters that are being deleted will show up in the list until all their resources are successfully removed by our operators. We’ll continue to iterate on this feature in an upcoming release.
New cluster creation form: The cluster creation form has changed slightly to better suit the different types of clusters that Happa can create now depending on the installation.
The Giant Swarm logo in the top left now redirects to the dashboard instead of our website.
The App Catalog has seen a number of performance improvements.
Version links on the cluster detail page are underlined to show that they are interactive clickable elements.
The worker node limit is now enforced on KVM installations. In other words, it is no longer possible to create failing clusters that will not fit on your installation.
Navigating between organizations and the organization list using the back button works correctly now.
Opening organizations in a new tab using ctrl-click works correctly now.
Deep linking to an App in the App Catalog has been fixed.
Ingress Port Mappings for KVM installations are back. They were accidentally removed from the cluster detail screen.
We now show a checkmark in the organization list when that organization is configured to use custom credentials.
We would also like to note that we’ve improved test coverage significantly and the overall development experience. We have also added exception notification to Happa. When an exception occurs, we will now be notified of it so that we can investigate and fix issues that were previously unreported.
It is now possible to upload the ‘secret values’ for an app from the App Detail modal on the ‘Apps’ tab of a cluster.
The maximum number of nodes allowed when creating a cluster has been increased from 99 to 999.
The styling of the validation messages on the cluster create screen is now more consistent.
Number picker buttons become disabled when increasing or decreasing would result in an invalid number.
On KVM installations, the minimum and default values for CPU Core and Memory size have been adjusted:
Fixes problems with logging in using a normal password.
incubator, test, community.