Oracle Cloud Infrastructure Documentation

Connecting to Worker Nodes Using SSH

If you provided a public SSH key when creating the node pool in a cluster, the public key is installed on all worker nodes in the cluster. On UNIX and UNIX-like platforms (including Solaris and Linux), you can then connect through SSH to the worker nodes using the ssh utility (an SSH client) to perform administrative tasks.

Note the following instructions assume the UNIX machine you use to connect to the worker node:

  • Has the ssh utility installed.
  • Has access to the SSH private key file paired with the SSH public key that was specified when the cluster was created.

Before you can connect to a worker node using SSH, you must define a security ingress rule in the security list for the worker node subnet to allow SSH access. The ingress rule must allow access to port 22 on worker nodes from source and any source port, as follows:

Type Source CIDR IP Protocol Source Port Range Dest. Port Range Type and Code Allows Purpose
Stateful TCP All 22 n/a TCP traffic for ports: 22 SSH Remote Login Protocol Enables SSH access.

To connect to a worker node through SSH from a UNIX machine using the ssh utility:

  1. Find out the public IP address of the worker node to which you want to connect. You can do this in a number of ways:

    • Using kubectl. If you haven't done so already, download the cluster's kubeconfig configuration file. If the file does not have the expected default name and location of $HOME/.kube/config, set the KUBECONFIG environment variable to point to the file. See Downloading a kubeconfig File to Enable Cluster Access. Then in a terminal window, enter kubectl get nodes to see the public IP addresses of worker nodes in node pools in the cluster.
    • Using the Console. In the Console, display the Cluster List page and then select the cluster to which the worker node belongs. Click Node Pools to see the public IP addresses of worker nodes in every node pool in the cluster.
    • Using the REST API. Use the ListNodePools operation to see the public IP addresses of worker nodes in a node pool.
  2. In the terminal window, enter ssh opc@<node_ip_address> to connect to the worker node, where <node_ip_address> is the IP address of the worker node that you made a note of earlier. For example, you might enter ssh opc@

    Note that if the SSH private key is not stored in the file or in the path that the ssh utility expects (for example, the ssh utility might expect the private key to be stored in ~/.ssh/id_rsa), you must explicitly specify the private key filename and location in one of two ways:

    • Use the -i option to specify the filename and location of the private key. For example, ssh -i ~/.ssh/my_keys/my_host_key_filename opc@
    • Add the private key filename and location to an SSH configuration file, either the client configuration file (~/.ssh/config) if it exists, or the system-wide client configuration file (/etc/ssh/ssh_config). For example, you might add the following:

      Host IdentityFile ~/.ssh/my_keys/my_host_key_filename

    For more about the ssh utility’s configuration file, enter man ssh_config

    Note also that permissions on the private key file must allow you read/write/execute access, but prevent other users from accessing the file. For example, to set appropriate permissions, you might enter chmod 600 ~/.ssh/my_keys/my_host_key_filename. If permissions are not set correctly and the private key file is accessible to other users, the ssh utility will simply ignore the private key file.