Skip to main content

Known Installation Issues

Metrics server does not start with untrusted Kube API certificates

Issue

Metrics server will not start when installing on a K8s cluster (like on Minikube or a cluster created with Kubeadm) with Kube API using self-signed certificates

Solution

Add extra args to the metrics-service by using the following values when installing Otomi with Helm chart:

apps:
metrics-server:
enabled: true
_rawValues:
extraArgs:
kubelet-preferred-address-types: InternalIP
kubelet-insecure-tls: true

Namespaces stuck in terminating state when uninstalling Otomi

Issue

When uninstalling Otomi using the helm unistall cmd, all Otomi namespaces get stuck in a terminating state.

Solution

The work around for now is to delete all namespaces using this cmd:

for ns in $(kubectl get ns --field-selector status.phase=Terminating -o jsonpath='{.items[*].metadata.name}'); do  kubectl get ns $ns -ojson | jq '.spec.finalizers = []' | kubectl replace --raw "/api/v1/namespaces/$ns/finalize" -f -; done

Installing Otomi with DNS fails due to failed authentication for Gitea

Issue

When installing Otomi with DNS, the installations fails with the following error:

otomi:cmd:commit:commitAndPush:error remote: Unauthorized
fatal: Authentication failed for 'https://gitea.otomi.example.com/otomi/values.git/'

Solution

Provide a custom password:

otomi:
adminPassword: yourpassword

Installing Otomi on Civo fails

There can be multiple issues when installing Otomi on Civo:

Issue

Some pods get stuck in a pending state during installation.

Solution

Otomi installs K8s applications like KeyCloak and Gitea. These apps require persistent storage. Civo has default quota set on resources, including volumes. If a pod stays in a pending state, it might be caused by a quota limit. In Civo dashboard, check the amount of volumes and remove all volumes that are not in use. Also check in the amount of volumes did not reach a quota limit and request an increase when needed.

Issue

When installing Otomi using Civo market place on a Linux Talos cluster the Otomi installation fails.

Solution

Installing Otomi on Civo Linux Talos is currently not supported. Try installing Otomi on Civo K3s.

Issue

Keycloak-01 pod is not able to start and the Keycloak logs show the following message: File "base/16385/PG_VERSION" does not contain valid data.

Solution

This issue is still under investigation. For now, first run the following cmd:

kubectl exec -it -n keycloak keycloak-db-1 -c postgres -- sh -c 'echo "15" > /var/lib/postgresql/data/pgdata/base/16385/PG_VERSION'

and then delete the keycloak-0 pod:

kubectl delete pod keycloak-0 -n keycloak

Keycloak should now start.