I just updated to the latest version of rancher v0.30.0 and all of my hosts are stuck in reconnecting and all thowing this error when i do a docker logs rancher-agent on the host.
INFO: Starting agent for 5B9E2A5355BDDF207386
INFO: Access Key: 5B9E2A5355BDDF207386
INFO: Config URL: http://rancher:8080/v1
INFO: Storage URL: http://rancher:8080/v1
INFO: API URL: http://rancher:8080/v1
INFO: IP: xxx.yyy.zzz.sss
INFO: Port:
INFO: Required Image: rancher/agent:v0.7.11
INFO: Current Image: rancher/agent:v0.7.11
INFO: Using image rancher/agent:v0.7.11
INFO: Downloading agent http://rancher:8080/v1/configcontent/configscripts
tar: Child returned status 1
tar: Error is not recoverable: exiting now
{“id”:“73503eb0-9a3a-44fc-b098-3788bfc9a32d”,“type”:“error”,“links”:{},“actions”:{},“status”:401,“code”:“Unauthorized”,“message”:“Unauthorized”,“detail”:null}
gzip: stdin: not in gzip format
tar: Child returned status 1
tar: Error is not recoverable: exiting now
That’s terrible. If you do a docker inspect rancher-agent you should see the agent has a bunch of environment variables set. If you take the environment set can you see if curl -u ${CATTLE_ACCESS_KEY}:${CATTLE_SECRET_KEY} ${CATTLE_URL} works?
Just an update, I upgraded my other rancher server this am and it was fine. The only difference was that the one I did last night was not v0.28 but the initial beta release (v0.25 i believe).
The server I did this am was on v0.28 and it started up fine and re-connected all the hosts. I’ve already rebuilt the other hosts as it was testing environment but I hope that helps.