Appendix A: Troubleshooting
ISSUE | ACTIONS |
How to find the installation ID? | See the Settings > License page on ICE Desktop |
What is the approximate time required to complete the successful install to plan the activity with network and system administrator? | Less than 30 minutes on a properly configured Kubernetes cluster |
When installing or using helm, the "Kubernetes cluster unreachable" error appears. | Make sure the Kubernetes cluster is accessible by running kubectl get nodes. Make sure the environment variable ¢KUBECONFIG is defined and pointing to a valid Kubernetes KUBECONFIG file, typically ¢{HOME}/.kube/config |
When installing using SSH, the session timed out and was disconnected before the install completed. | The ICE Server installation may continue to run when your session is disconnected. Simply resume installation from where you left off. |
How to check if the ICE Server charts and add-ons are installed? | Run helm ls -A then kubectl get pods -A to look for pods that failed to start. Consult technical support if there is any pod that shows large number of restarts. |
How to view the current version. | On ICE Desktop, click Help > Build Info. |
ICE Server superuser password is lost. How to reset it? | Use another administrator account to reset the password |
How to request a license? | Contact ICE License Support with installation ID. |
License file upload failed | Make sure the license file received from ICE License Support is saved as-is, without any modification. Make sure the installation ID in the license file matches what is displayed on the ICE Desktop license page. |
How to increase the licensed feature counts? | Request a new license from ICE Sales Support |
Is a new license necessary if the product is reinstalled? | Yes. Any new installation (including reinstallation on the same Kubernetes cluster) will require a new license. |
The hosting VM is rebooted. Is manual restart of the ICE Server necessary? | No. On host reboot, the ICE Server restarts automatically. You can run watch kubectl get pods -A to monitor pod restart status. The pods may take a few minutes to up to 15 minutes (on slower system) to complete restart. Typically, restarting the host VM is not recommended, as it rarely resolves any pod issue. |
What is the approximate time required to complete the successful upgrade to plan the activity with network and system administrator? | Upgrade typically only requires a brief, transient outage of less than one minute. Active users typically do not need to log out during the upgrade process. |
watch kubectl get pods -A is showing spordic etcdserver timeout errors | Your hosting hardware's storage devices may be too slow. Review disk I/O latency of your hosting hardware, upgrade storage devices as needed |
Is it possible to change IP address and/or hostname after ICE Server is installed? | After the cluster is installed, changing IP address and/or hostnames is not recommended |