Common deployment errors

This topic lists some common errors that you might encounter when you are deploying the Cloud App Management product. The error messages that are displayed, the reasons for these errors occurring, and the solutions for them are included in the following information.

Problem - Certificate error

Error message such as:

could not read x509 key pair (cert: "/root/.helm/cert.pem", key: "/root/.helm/key.pem"): 
can't load key pair from cert /root/.helm/cert.pem
and key /root/.helm/key.pem: open /root/.helm/cert.pem: no such file or directory

Symptoms

This error occurs when you try to run any Helm command to connect the Helm server. The certificate key files are not installed correctly.

Solution

Find the pem keyfiles and copy them to the /root/.helm default location. You can manually copy over these keyfiles to the ~/.helm/ default directory by issuing the following command:

 cp ~/.kube/mycluster/*.pem ~/.helm/

Problem - Helm cannot connect error

The error message that is displayed depends on the Helm command that you are running.

Error: cannot connect to Tiller

Error: transport is closing

Symptoms

This error occurs when you try to run a helm command and you did not add the --tls parameter with this command.

Solution

Add --tls to the helm command that you are running. For example:

helm install --tls

Problem - kubectl or Helm errors because authorization has expired

For example, error messages such as the following can be displayed

kubectl get pods
Error: the server doesn't have a resource type "pods"

Symptoms

These type of errors occur when you are not actively working in your IBM® Cloud Private cluster from the IBM Cloud Private CLI for a long time such as a few hours. Your authorization has expired.

Solution

You must log in to the IBM Cloud Private cluster again by issuing the following command:

cloudctl login -a https://mycluster.icp:8443 --skip-ssl-validation -u admin

Where mycluster.icp is the default IBM Cloud Private cluster name. This name might be something different if you entered your own cluster name when you were deploying IBM Cloud Private.

Problem - Deployment already exists error

Error message such as:

deployment my_deployment_name already exist

where my_deployment_name is the deployment name.

Symptoms

This error occurs when the previous deployment with the same name as the deployment you are attempting to install now didn't delete successfully.

Solution

You must find this deployment and delete it manually. For example, to find the deployment, issue the following command:


kubectl get deployment

Next, you must manually delete the deployment by issuing the following command:


kubectl delete deploy my_deployment_name