The version of Rancher is located in the footer of our site. If you click on the version, you’ll be able to get specific versions of our other components.
Read more about how to install Rancher server behind a proxy.
docker logs on the Rancher server container will provide a set of the basic logs. To get more detailed logs, you can exec into the rancher server container and look at the log files.
# Exec into the server container $ docker exec -it <container_id> bash # Navigate to the location of the cattle logs $ cd /var/lib/cattle/logs/ $ cat cattle-debug.log
Inside this folder, there will be
cattle-error.log. If you have been using Rancher server for many days, you will find a log file for each day as we create a new file for each day.
Here’s the command to copy the Rancher server logs from the container to the host.
$ docker cp <container_id>:/var/lib/cattle/logs /local/path
If the IP of Rancher server has been changed, you will need to re-attach the hosts with the updated information.
In Rancher, go to Admin -> Settings and update the Host Registration with the updated URL for Rancher server. Please note that it must include the exposed port that you started Rancher server with. By default, we have recommended using port
8080 in our installation instructions.
After the Host Registration has been updated, go to Infrastructure -> Add Hosts -> Custom. The
docker run command to add Rancher agents will have been updated with the new information. Using the updated command, run the command on all the hosts inside your Rancher server environments.
Go-machine-service is a micro-service that connects to the Rancher API server via a websocket connection. If it fails to connect, it restarts and tries again.
If you are running in Single Node setup, it will connect to the Rancher API server using the URL you set for Host Registration. Verify that the Host Registration URL can be reached from inside the rancher-sever container.
$ docker exec -it <rancher-server_container_id> bash # Inside the rancher-server container $ curl -i <Host Registration URL you set in UI>/v1
You should get a json response. If authentication is turned on, the response code should be 401. If authentication is not turned on, the response code should be 200.
If you are running in Multi Node setup, then you started go-machine-service in its own container and explicitly provided a URL, access key, and secret key as environment variables upon startup. Verify that the Rancher API server can be reached using those variables. Verify the connectivity by logging into the go-machine-service container and making a curl command using the parameters your provided to the container:
$ docker exec -it <go-machine-service_container_id> bash # Inside the go-machine-service container $ curl -i -u '<value of CATTLE_ACCESS_KEY>:<value of CATTLE_SECRET_KEY>' <value of CATTLE_URL>
You should get a json response back and a 200 response code.
If the curl command fails, then you have a connectivity issue between go-machine-service and the Rancher API server.
If the curl command does not fail, the problem could be related to the fact that go-machine-service is attempting to establish a websocket connection, not a normal http connection. If you have a proxy or load balancer between go-machine-service and your Rancher API server, verify that it is configured to allow websocket connections.
Most likely there are some tasks that are stuck running for some reason. If you are able to look at the Admin -> Processes tab in the UI, you’d be able to see what is stuck in
Running. If these tasks have been running (and failing) for a long time, Rancher ends up using too much memory to track the tasks. Essentially creates a running out of memory situation for Rancher server.
In order to get your server into a responsive state, you’ll need to add more memory. Typically, 4GB is more than sufficient.
If you run the Rancher server command again, just add an additional option,
-e JAVA_OPTS="-Xmx4096m" to the command.
$ docker run -d -p 8080:8080 --restart=always -e JAVA_OPTS="-Xmx4096m" rancher/server
Depending on how the MySQL database is setup, you may need to do an upgrade to add the additional command.
If you were unable to see the Admin -> Processes tab due to the lack of memory, after starting Rancher server again with more memory, you should be able to see the tab and start troubleshooting the processes that have been running the longest.
Rancher server automatically cleans up a couple of database tables to prevent the database from increasing too quickly. If you are noticing that these tables are not being cleaned up quick enough for you, please feel free to update the default settings using our API.
By default, any records in the
service_event tables are deleted if they were created 2 weeks ago. The setting in the API is listed in seconds (
1209600). The setting in the API is
By default, any records in the
process_instance table are deleted if they were created 1 day ago. The setting in the API is listed in seconds (
86400). The setting in the API is
To update these settings in your API, navigate to the
http://<rancher-server-ip>:8080/v1/settings page. Search for the setting you want to update and click on the link in the
links -> self to navigate to the setting. Click on
Edit on the side page to change the
value. Remember, the value is in seconds.
If you are starting Rancher and it freezes forever, there might be a liquibase database lock. On startup, liquibase does a schema migration. There is a race condition where it might leave a lock entry, which will prevent subsequent boots.
If you have just upgraded and in the Rancher server logs, there can be a log lock on the MySQL database that has not been released.
....liquibase.exception.LockException: Could not acquire change log lock. Currently locked by <container_ID>
Note: Please do not release the database lock unless you have the above exception regarding the log lock. If your upgrading is taking a long time due to data migration, you may hit other migration issues if you try to release the database lock.
If you had created the data container for Rancher server per the upgrading documentation, you’ll need to
exec into the
rancher-data container to update the
DATABASECHANGELOGLOCK table and remove the lock. If you hadn’t created the data container, you can
exec into the container that has your database.
$ sudo docker exec -it <container_id> mysql
Once you are in MySQL database, you’ll need to access the
mysql> use cattle; # Check that there is a lock in the table mysql> select * from DATABASECHANGELOGLOCK; # Update to remove the lock by the container mysql> update DATABASECHANGELOGLOCK set LOCKED="", LOCKGRANTED=null, LOCKEDBY=null where ID=1; # Check that the lock has been removed mysql> select * from DATABASECHANGELOGLOCK; +----+--------+-------------+----------+ | ID | LOCKED | LOCKGRANTED | LOCKEDBY | +----+--------+-------------+----------+ | 1 | | NULL | NULL | +----+--------+-------------+----------+ 1 row in set (0.00 sec)
When launching a HA setup of Rancher, you can add port
18080 to the external load balancer to view progress of the HA setup. In the Stacks -> Management stack, all the services required for the HA setup are being launched. When all the services are in
Active state, the Rancher UI will be available.