By clicking the selected project or K8s View you are redirected to the Kubernetes Servers. Here you can see all K8s servers for the project with their description.
Project Info #
Under Servers title is a brief description of the project – such as Project Name (with locked/unlocked image), Project Status, Cloud Type, Kubernetes Version, Access Profile, Cloud Credentials, Kubernetes Profile, Alerting Profile, Policy Profile, Access IP Address (if you use this address to ssh connect, please do not use user ubuntu, it’s in use by Taikun for managing the cluster) and Kubernetes Health.
You can also see here ETC=Estimated Time to Complete. It is approx time (in minutes) until the cluster will be completed.
Every Server is described by ID, Server Name, IP Address, Flavor, CPU/RAM/Disk Size, Role, Status, Kubernetes Health and Creation Time. If you expand the table, you can see the last modification made (Created By, Last Modified, Last Modified By).
Server status can be:
- Pending Delete
- Pending Upgrade
Upgrade your version of Kubespray to the latest one. The button is disabled, if your Kubernetes are up to date. Also if you checked Auto Upgrade during creation.
Sends the changes to the repository. Once the cluster is committed you will see ETC in project info.
When the server/s are Failed or there is some other problem in the cluster, use repair button.
If the project is locked ? which you can see under Project Info, you can only preview some pages (e.g. Kube Info). You can’t make any changes.
Add Server #
To create a new server click button and fill all the fields. You, as user, can’t delete servers – think twice which and how many servers you want to create.
Server Name – only alphanumeric characters and dash are allowed, 1-30 characters
Letters must be lowercase!
Disk size – should be at least 30GB
Role – you are able to choose from several roles for your servers, which are set according to the Kubernetes settings
Number of Servers – set number for kubeworker or kubemaster, add odd number of masters (min. 3 for a highly available cluster)
The change MUST be committed.
Flavor – choose from the list of offered flavors (e.g. n0.large)
Recommendation for sizing:
* bastion recommended 2 vCPU + 2GB of RAM
* masters recommended 4 vCPU + 8GB of RAM
Kubernetes Node Labels – label nodes where you want to sent or restrict pods; you can add more labels with add label button, for more info see Kubernetes docs.
Function buttons #
Under Add Server are buttons with specific features or more detailed information.
Add a new kubernetes configuration for your profile and project.
Kubeconfig Name – choose a name for your kubeconfig (3-230 characters)
- cluster-admin – perform any action on any resource, ClusterRoleBinding – gives full control over every resource in the cluster and in all namespaces (or in very resource in the role binding’s namespace – RoleBinding)
- admin – RoleBinding – allows read/write access to most resources in a namespace, does not allow write access to resource quota or to the namespace itself
- edit – allows read/write access to most objects in a namespace, does not allow viewing or modifying roles or role bindings, allows accessing Secrets and running Pods as any ServiceAccount in the namespace
- view – see most objects in a namespace, does not allow viewing roles or role bindings, does not allow viewing Secrets
For more info, see kubernetes documentation.
Personal Kubeconfig – kubeconfig can be used only by you
You can see all project’s configurations in the table with its ID, Name, User Name, User Role, Project, Accessible for all, Konfig Role Name, Created By and Actions.
This .yaml file can be download and use to organize information about clusters, users, namespaces, and authentication mechanisms.
Delete your kube config if it is no longer needed.
K8s Info #
If Kubernetes is active, Kube Info button will take you to the Kubernetes configuration. For more see Projects – Kubernetes.
You are redirected to Events, where you can see all Kubernetes changes made in the project. To preview details for more information to each action usebutton. A green strip indicates a successful action, a red strip indicates a failed action. Use Clear events for deleting all the events.
You can sort Events by:
- Search field
- Filling Start and End Date
- Tick Only failed to filter failed actions
Preview Kubernetes cluster logs to Grafana. Logs button is disabled if Monitoring is disabled.
Write your query and use Start date and End Date for sorting. You can also expand every message – red is added action, no color is other.
First thing when you access Alerts are Firing Alerts. This section is refreshed every 5 minutes, but you can also use the refresh button to see the most updated data.
To see all alerts, use upper right Show All Alerts button. As seen above, firing alerts are marked with red color. For each alert you can see details and use a link that will redirected you to Metrics with the query already written.
Alerts are accessible only if Monitoring is enabled and the project is not empty.
The index number at Alerts shows the number of firing alerts. When the firing alerts are resolved, the number disappears.
Firing alerts also work from the real-time notifications bell in the header. Red ones are indicate alert, green ones means that the alert is resolved.
The notification contains a brief message on a specific project and bellow the message is time of the change. After clicking Show Project, you can access the project in which the alert is.
As the name suggests: Mark as Read and Dismiss all.
Write a query, search Prometheus Metrics and preview the value needed. Modify Step or Date.
Switch between Console and Graph for better results.
Project Dashboard #
Dashboard is accessible only if cluster is created and monitoring is enabled.