Skip to content

Installation Troubleshooting

This page summarizes common issues with the installer and provides troubleshooting solutions to help users quickly resolve problems encountered during installation and operation.

Troubleshooting DCE 5.0 Platform Interface Unavailability with diag.sh Script

Since installer version v0.12.0, a new diag.sh script has been added to facilitate troubleshooting when the DCE 5.0 platform interface is unavailable.

Run the command:

./offline/diag.sh

Example of execution result:

FAQ1

Failure to Start kubelet Service After Kind Container Restart

After restarting the Kind container, the kubelet service fails to start and reports an error:

failed to initialize top level QOS containers: root container [kubelet kubepods] doesn't exist

Solution:

  • Solution 1: Restart by executing the command podman restart [kind] --time 120, and do not interrupt this task with ctrl+c during the execution process.

  • Solution 2: Enter the Kind container using podman exec and Run the following command:

    for i in $(systemctl list-unit-files --no-legend --no-pager -l | grep --color=never -o .*.slice | grep kubepod);
    do systemctl stop $i;
    done
    

Podman fails to create containers after disabling IPv6

Error message:

ERROR: failed to create cluster: command "podman run --name kind-control-plane...  

Solutions

Re-enable IPv6 or change bootstrapping node base to Docker.

Podman-related issues: https://github.com/containers/podman/issues/13388

Redis get stuck when reinstalling DCE 5.0 in kind cluster

Issue: The Redis Pod remains in a 0/4 running state for a long time with the error message primary ClusterIP can not unset.

  1. Delete the rfs-mcamel-common-redis service in the mcamel-system namespace:

    kubectl delete svc rfs-mcamel-common-redis -n mcamel-system
    
  2. Then, retry the installation command.

When using Metallb, the VIP access is blocked and cannot log into DCE

  1. Check whether the VIP is in the same network segment as the host. In Metallb L2 mode, they should be in the same network segment.

  2. If this error occurs after you added a new NIC to the control node in the global cluster, you need to manually declare and configure L2Advertisement.

    Refer to related Metallb issues

Community package: fluent-bit installation failed

Error:

DaemonSet is not ready: insight-system/insight-agent-fluent-bit. 0 out of 2 expected pods are ready

Solutions:

Check if the following key information appears in the Pod log:

[warn] [net] getaddrinfo(host='mcamel-common-es-cluster-masters-es-http.mcamel-system.svc.cluster.local',errt11):Could not contact DNS servers

If yes, it is a known bug of fluent-bit bug. Refer to: https://github.com/aws/aws-for-fluent-bit/issues/233

Error reported during CentOS 7.6 installation

Error:

FAQ1

Solutions:

Run modprobe br_netfilter on each node where the global service cluster is installed, and wait until br_netfilter is loaded.

CentOS environment preparation issues

Running yum install docker reports an error:

Failed to set locale, defaulting to C.UTF-8
CentOS Linux 8 - AppStream                                                                    93  B/s |  38  B     00:00    
Error: Failed to download metadata for repo 'appstream': Cannot prepare internal mirrorlist: No URLs in mirrorlist

You can try the following methods to resolve the issue:

  • Install glibc-langpack-en package:

    sudo yum install -y glibc-langpack-en
    
  • If the issue persists, try the following:

    sed -i 's/mirrorlist/#mirrorlist/g' /etc/yum.repos.d/CentOS-*
    sed -i 's|#baseurl=http://mirror.centos.org|baseurl=http://vault.centos.org|g' /etc/yum.repos.d/CentOS-*
    sudo yum update -y
    

Unable to Restart kind Cluster Properly After Bootstrap Node Reboot

After rebooting the bootstrap node, the kind cluster may fail to restart properly because the kind cluster was not set to start automatically on the openEuler 22.03 LTS SP2 operating system during deployment.

To resolve this issue, execute the following command to restart:

podman restart $(podman ps | grep installer-control-plane | awk '{print $1}') 

Note

If the above scenario occurs in other environments, you can also execute the same command for restarting.

Missing ip6tables When Deploying Ubuntu 20.04 as the Bootstrap Machine

When deploying Ubuntu 20.04 as the bootstrap machine, the absence of ip6tables can cause errors during the deployment process.

Refer to the Podman known issue.

Temporary solution: Manually install iptables, refer to Install and Use iptables on Ubuntu 22.04.

Comments