Fault description: The
cluster deployment service keeps reporting an x509 error, causing the pod to fail to start
Solution steps:
1. First check the certificate and check whether all the certificates have problems.
Use md5sum to compare the md5 values of the certificates to see if they are consistent. The final result is that the certificates are all consistent, no problem.
2. Check the endpoints of the cluster and use the kubectl get ep command to find out that there is a problem with the cluster. We have 3 masters in total, but 2 are found. There are still problems with these two. It may be that the colleague did not modify the ep after changing the master to another host. As a result, it is now found that the ep value is one more host than .5, and the IP of the 61 and 63 hosts are missing.
3. Modify the yaml file of ep, remove the extra useless .5 host, and add some other 2 master hosts.
4. The modification is successful and the problem is solved.