Unable to log into Rancher Gui after reboot

Rancher
1

Hello, I just spent a few days rebuilding my rancher cluster. I had access to the Rancher GUI. I had to restart the physical server and now I can’t connect to the GUI. The rancher/rancher-agent:v2.3.1 keeps restarting over and over and the logs say
[root@ca819-elastic01 containers]# tail -f

“log”:“INFO: Using resolv.conf: search cert.scrub.com nameserver 192.168.140.10\n”,“stream”:“stderr”,“time”:“2023-05-09T23:11:52.170719473Z”}
{“log”:“ERROR: https://192.168.29.54:8443/ping is not accessible (Failed to connect to 192.168.29.54 port 8443: Connection refused)\n”,“stream”:“stderr”,“time”:“2023-05-09T23:11:52.203669229Z”}

here was my original installation command
docker run -d --restart=unless-stopped -p 8080:80 -p 8443:443 --privileged rancher/rancher:v2.3.1

could i install the rancher client without wiping out my cluster? I’m not sure how to fix this issue other than guessing at trying something.

3

here is the rest of the log message since only 2 links can be put into a ticket at a time

cb084998bd1cbcfa772bc513a3d7be868b87fec0da954c1d788995072212e396 --no-register --only-write-certs\n",“stream”:“stderr”,“time”:“2023-05-09T23:12:52.548637034Z”}
{“log”:“INFO: Environment: CATTLE_ADDRESS=192.168.29.54 CATTLE_AGENT_CONNECT=true CATTLE_INTERNAL_ADDRESS= CATTLE_NODE_NAME=ca819-elastic01 CATTLE_SERVER=https://192.168.29.54:8443 CATTLE_TOKEN=REDACTED CATTLE_WRITE_CERT_ONLY=true\n”,“stream”:“stderr”,“time”:“2023-05-09T23:12:52.573724965Z”}

4

Evidently, I do not have the docker proxy running on my new physical server running the new cluster via ss command. However, on a working cluster it shows
tcp LISTEN 0 128 [::]:8443 [::]:* users:((“docker-proxy”,pid=31821,fd=4)). On the new cluster it is not running. I’m not sure how to get it running without breaking the cluster as it is working, i just can’t connect with gui.
the install command of the rancher gui container was
docker run -d --restart=unless-stopped -p 8080:80 -p 8443:443 --privileged rancher/rancher:v2.3.1
so i dont understand why 8443 is not a port listening in my centos

5

figured it out, i ran docker ps -a, found the rancher/rancher container Id and docker start and all works, what threw me off is the container ID was differerent than the container ID after the brand new install of rancher/rancher.

6

The error message indicates that the Rancher agent is unable to connect to the specified URL, resulting in a connection refusal. You can verify network connectivity, ensure the rancher server is running, and restart the rancher agent container. Installing the Rancher on your local machine should not affect your existing cluster.