farosctl COMMAND [OPTIONS]
The farosctl command is responsible for configuring the bastion node/router, configuring the OpenShift cluster, and performing some day 2 operations on the cluster.
The following commands are available.
Print the directory used for storing the cluster configuration.
farosctl extract SOURCE DEST
Search for the
SOURCE file in the binary and export it to
on the host. This is useful for extracting new versions of the oc or
farosctl extract oc ~/bin/oc
farosctl [SUBCOMMAND] help
Print help documentation.
Launch a farosctl managment shell. Useful for debugging issues.
Update faros to the newest version.
farosctl apply [SUBCOMMAND] [FLAGS]
Apply commands are idempotent and safe to run at any time, even after the cluster has been created.
[DEFAULT] Apply all configuration
Apply DNS and DHCP reservations
Apply the network routing and firewall configuration
farosctl apply host-records
farosctl config [SUBCOMMAND]
Launch the configuration TUIs to set the parameters that control the final cluster behavior.
[DEFAULT] Configure only the main cluster settings
Configure only the main cluster settings
Configure proxy settings required to reach the internet
farosctl create SUBCOMMAND [FLAGS]
Create commands are not to be considered idempotent and should (generally) not be run after the cluster has been created. These commands are potentially destructive actions that are taken in order to create the initial cluster.
Perform the final step of install OpenShift
Create the internal sources for the RHCOS images and Ignition files
Create the cluster load balancer
Create any virtual machines required for install (eg: Bootstrap node)
farosctl create install-repos
farosctl deploy SUBCOMMAND [FLAGS]
Deploy commands are generally run after the cluster has been created and are responsible for adding services to the OpenShift cluster and performing post-install configuration of the cluster.
Deploy OpenShift Container Storage
[EXPERIMENTAL] Deploy the cluster load balancer as a service on the cluster
Deploy NVIDIA operator to install drivers on cluster nodes.
Deploy an example Open Data Hub workflow
Deploy entitlement certificates to the cluster nodes to allow the cluster to build UBI images.
Wipe the filesystem of select drives on select nodes
farosctl deploy container-storage
farosctl destroy SUBCOMMAND [FLAGS]
Destroy various resouces that are created during the Faros deployment. These actions are extremely destructive and should be done with care.
[DEFAULT] Run all of the destroy steps in order
Destroy the deployed OpenShift cluster and wipe the node disks
Clean all local content repositories with OS images and configs.
Remove the load balancer the fronts the OpenShift cluster
farosctl destroy cluster
farosctl get SUBCOMMAND [FLAGS]
These commands return information about the cluster for convenience.
Show the overall cluster status and health
Return the URLs and passwords required to login to the cluster
Display a table of all IP addresses on the cluster network that are known to the faros utility
Returns the oc command used to login to the cluster
Returns the DNS records that are required external to the cluster
farosctl get credentials
farosctl identify HOST_LIST
Blink the identification light on the hosts specified via a comma seperated list.
farosctl identify node-0
farosctl install-plan SUBCOMMAND [FLAGS]
Installation plans are a series of farosctl commands that are commonly run together to accomplish a larger task.
Apply configuration, create machines, and do everything else required to make a cluster from scratch
Ensure the cluster nodes are properly entitled and then configure them to support NVIDIA GPUs.
farosctl install-plan cluster
farosctl oc [OC FLAGS AND ARGUMENTS]
This command runs oc as the kubeadmin user with full cluster rights. This is provided for convenience and some oc commands may not work properly (oc patch, for example). It is recomended to login to the cluster natively from the command line instead of relying on farosctl os.
farosctl oc get nodes
poweroff / shutdown¶
Safely shutdown the cluster.
force-poweroff / force-shutdown¶
Poweroff the cluster without verifying the health of the cluster operators.
poweron / startup¶
Bring the cluster up.
farosctl ssh NODE_NAME
Use SSH to make a connection to any of the nodes in the cluster. Use the hostnames for the nodes, not the FQDN.
farosctl ssh node-0
Print the version of the current farosctl code.
farosctl wait-for SUBCOMMAND [FLAGS]
Wait for commands are designed to block the prompt until a condition is met.
Wait for the cluster node firmware configuration to be manually verified by the user.
Wait for the management interfaces on the cluster nodes to come online with the expected IP addresses
farosctl wait-for management-interfaces
If it is desired to lock farosctl to a specific release, this can be easily done by creating a symbolic link to the farosctl executable with the desired version appended to the executable name.
For example, in order to pin to version 0.5.6:
cd ~/bin ln -s farosctl farosctl.0.5.6 farosctl.0.5.6 version
Even if an update is run, this link will always run the 0.5.6 release.
The same technique can also be used to run the latest development build of the farosctl.
cd ~/bin ln -s farosctl farosctl.dev farosctl.dev version