Skip to main content

Troubleshooting & FAQ

When we try to login to the Appcircle server, we see too many redirects error from browser​

This error usually happens when the pods can't resolve some of the Appcircle server domains.

For the solution, please make sure that the domain name server of the worker nodes of the Kubernetes cluster can resolve the Appcircle server domain names.

When we deploy the Helm chart, the appcircle-server-webeventredis-master-0 pod is stuck in CrashLoopBackOff state​

This error usually happens when you select a non-valid Appcircle CA Certificate File while creating the configuration file. Please make sure that the certificate you choose is the root certificate of the full-chain certificate.

tip

If you created the SSL/TLS certificate with LetsEncrypt, you should know that the fullchain.pem file doesn't include the root CA certificate by default.

To fix the problem, you can edit the values.yaml file and upgrade the Helm chart.

helm upgrade appcircle-server appcircle/appcircle \
  --timeout 1200s \
  -n appcircle \
  -f values.yaml
caution

The stateful pods won't be recreated from a error state. This is known issue of Kubernetes.

You should delete the pods manually to fix this problem. The new updated pods will be created automatically. You can use the example commands below to delete the pods:

kubectl delete pods appcircle-server-webeventredis-master-0 -n appcircle && \
kubectl delete pods appcircle-server-webeventredis-replicas-0 -n appcircle && \
kubectl delete pods appcircle-server-webeventredis-replicas-1 -n appcircle

What should we do if the deployment hasn't been completed and timed out?​

If the deployment hasn't completed and timed out after 1200 seconds:

  • Low Network Bandwidth or Insufficient Processing Power: If the timeout occurred due to low network bandwidth or insufficient processing power, you can re-run the Helm deployment command as it is idempotent.

  • Configuration Issues: If the timeout was caused by a configuration problem, you will need to troubleshoot the issue. Review your configuration settings and logs to identify and resolve any errors before attempting the deployment again.

tip

If you face a timeout due to configuration problems, it is better to re-install Appcircle freshly. Refer to the Uninstalling Appcircle section for detailed instructions on how to uninstall and clean up the existing deployment before starting anew.

What should we do if we use an Ingress controller other then Nginx Ingress?​

If you are using an ingress controller other than the Nginx Ingress controller, you should add an additional annotation to the resource ingress.

By default, Appcircle adds the upstream vhost annotation. You should add alternative annotation for your ingress controller.

For example, if you are using HAProxy as an Ingress controller, you should add the annotation to the values.yaml of the Appcircle Helm chart:

caution

appcircle-server below should be the Helm chart release name. In the installation document, we use appcircle-server for the release name. You should change it if you changed the release name.

resource:
  ingress:
    annotations: 
      haproxy.org/set-host: "appcircle-server-minio:9000"