Occasionally, you might see errors or failures as the addons you are deploying are enabled when you run the konvoy up
command.
In most cases, these errors indicate that one or more addons did not deploy properly but do not specify the reason for the deployment failure.
The topics in this section cover some of the basic troubleshooting steps you can take to debug failures with addons.
Check the status of addon deployment
When addons are deployed at the end of konvoy up
or konvoy deploy
command execution, the konvoy command-line program emit statuses for each addon that is deployed showing whether they succeeded or failed:
STAGE [Deploying Enabled Addons]
konvoyconfig [OK]
dashboard [OK]
external-dns [OK]
reloader [OK]
opsportal [OK]
cert-manager [OK]
gatekeeper [OK]
defaultstorageclass-protection [OK]
traefik [OK]
awsebscsiprovisioner [OK]
dex [OK]
kube-oidc-proxy [OK]
traefik-forward-auth [OK]
prometheus [OK]
dex-k8s-authenticator [OK]
prometheusadapter [OK]
velero [OK]
kommander [OK]
elasticsearch [OK]
elasticsearch-curator [OK]
elasticsearchexporter [OK]
fluentbit [OK]
kibana [OK]
The above shows an addon deployment where no issues encountered, however alternatively you could see something like:
traefik [ERROR]
Error: error installing: the following 1 addon failed to deploy: [ traefik ]
If possible Konvoy will emit information about the failure directly in this output underneath the error, which can be helpful for diagnosing the root cause and making the changes necessary to accommodate before running konvoy deploy addons
again.
Many of the addons (including traefik
) are deployed using helm.
It may be helpful to start troubleshooting with:
helm list
This will provide a status output for all helm releases.
It can also be helpful to check the deployments or pods for any given addon.
For instance, if you had a problem with Traefik as seen above, the first step could be to inspect the traefik deployment:
kubectl -n kubeaddons get deployment traefik-kubeaddons
NAME READY UP-TO-DATE AVAILABLE AGE
traefik-kubeaddons 0/2 0 0 10m
This indicates a problem with deploying the pods, which you can investigate by looking through the deployment information:
kubectl -n kubeaddons describe deployment traefik-kubeaddons
The events and information there may point you to the cause of the problem, or if the pod simply failed, you can view the available traefik pods:
kubectl -n kubeaddons get pods | grep traefik
Collect the pod name (e.g. traefik-kubeaddons-<id>
) here and then investigate further by pulling the logs for each pod:
kubectl -n kubeaddons logs traefik-kubeaddons-<id>
NOTE: It can be the case that only [ERROR]
is received with no information about the failure.
In this case, report an issue through the regular channels.