CLI tool for Platform9 management.
- CPUs: Minimum 2 CPUs needed on host
- RAM: 12 GB
- Disk: At least 30 GB of total disk space and 15 GB of free space is needed on host
- Sudo access to the user
- OS(Supported) :
- Ubuntu (16.04,18.04,20.04)
- RHEL/Centos (7.x)
The CLI allows configuration where all HTTPS requests can be routed through a proxy. See the Configuration
section to see how to configure the proxy URL.
The CLI can be run in a non-interactive mode with flag --no-prompt
. Using this disables all user prompts. If required flags are not passed to a sub-command or in case of any error, the CLI returns with a non zero code.
- Downloading the CLI
bash <(curl -sL https://pmkft-assets.s3-us-west-1.amazonaws.com/pf9ctl_setup)
- Help
#pf9ctl --help
CLI tool for Platform9 management.
Platform9 Managed Kubernetes cluster operations. Read more at
http://pf9.io/cli_clhelp.
Usage:
pf9ctl [command]
Available Commands:
attach-node Attaches a node to the Kubernetes cluster
authorize-node Authorizes this node with PMK control plane
bootstrap Creates a single-node Kubernetes cluster using the current node
bundle Gathers the support bundle and uploads it to S3
check-amazon-provider Checks if the user has Amazon cloud permissions
check-azure-provider Checks if the user has Azure cloud permissions
check-google-provider Checks if the user has Google cloud permissions
check-node Checks prerequisites on a node to use with PMK
config Creates or get the config
deauthorize-node Deauthorizes this node from the PMK control plane
decommission-node Decommissions this node from the PMK control plane
delete-cluster Deletes the cluster
detach-node Detaches a node from a Kubernetes cluster
help Help about any command
prep-node Sets up prerequisites & prepares a node to use with PMK
upgrade Checks for a new version of the CLI
version Prints current version of CLI being used
Flags:
-h, --help help for pf9ctl
--log-dir string path to save logs
--no-prompt disable all user prompts
--verbose print verbose logs
Use "pf9ctl [command] --help" for more information about a command.
-
Version
This command is used to get the current version of the CLI
#pf9ctl version
pf9ctl version: v1.16
-
Upgrading
This command is used upgrade the CLI to its newest version if there is one
#pf9ctl upgrade
You already have the newest version
#pf9ctl upgrade
New version found. Please upgrade to the newest version
Do you want to upgrade? (y/n): y
Downloading the CLI
Installing the CLI
Successfully updated.
#pf9ctl upgrade --skip-check
New version found. Please upgrade to the newest version
Downloading the CLI
Installing the CLI
Successfully updated.
-
Configuration
This is used to setup or get the control-plane configuration. It includes the DU FQDN , username, region and the tenant(service).
#pf9ctl config
Create or get PF9 controller config used by this CLI
Usage:
pf9ctl config [command]
Available Commands:
get Print stored config
set Create a new config
Flags:
-h, --help help for config
Global Flags:
--no-prompt disable all user prompts
--verbose print verbose logs
Use "pf9ctl config [command] --help" for more information about a command.
#pf9ctl config set --help
We can set config through prompt or with flags.
Create a new config that can be used to query Platform9 controller
Usage:
pf9ctl config set [flags]
Flags:
-u, --account-url string sets account_url
-h, --help help for set
--mfa string set MFA token
-p, --password string sets password (use 'single quotes' to pass password)
-l, --proxy-url string sets proxy URL, can be specified as [<protocol>][<username>:<password>@]<host>:<port>
-r, --region string sets region
-t, --tenant string sets tenant
-e, --username string sets username
Global Flags:
--log-dir string path to save logs
--no-prompt disable all user prompts
--verbose print verbose logs
-
Check Node
This command will perform the prerequisite check before a node can be added to the cluster. It checks for the following:
-
Mandatory Checks
- Existing pf9 packages
- Installing missing packages
- Sudo Access Check
- Ports Check
- OS Check
- Existing Kubernetes Cluster Check
- Check execute permissions on /tmp folder
- Disabling swap and removing swap in fstab
-
Optional Checks
- Removal of existing pf9ctl(Python Based CLI)
- Resources(CPU,Disk,Memory check)
-
#pf9ctl check-node --help
Check if a node satisfies prerequisites to be ready to be added to a Kubernetes cluster. Read more
at https://platform9.com/blog/support/managed-container-cloud-requirements-checklist/
Usage:
pf9ctl check-node [flags]
Flags:
-h, --help help for check-node
-i, --ip strings IP address of host to be prepared
--mfa string MFA token
-p, --password string ssh password for the nodes (use 'single quotes' to pass password)
-s, --ssh-key string ssh key file for connecting to the nodes
-e, --sudo-pass string sudo password for user on remote host
-u, --user string ssh username for the nodes
Global Flags:
--log-dir string path to save logs
--no-prompt disable all user prompts
--verbose print verbose logs
Check-Node(Local)
#pf9ctl check-node
✓ Loaded Config Successfully
✓ Removal of existing CLI
✓ Existing Platform9 Packages Check
✓ Required OS Packages Check
✓ SudoCheck
✓ CPUCheck
✓ DiskCheck
✓ MemoryCheck
✓ PortCheck
✓ Existing Kubernetes Cluster Check
✓ Completed Pre-Requisite Checks successfully
Check-Node(Remote)
#pf9ctl check-node -i 10.128.241.203 -u centos
✓ Loaded Config Successfully
You can choose either password or sshKey
Enter 1 for password and 2 for sshKey
Enter Option : 1
Enter password for remote host:
✓ Removal of existing CLI
✓ Existing Platform9 Packages Check
✓ Required OS Packages Check
✓ SudoCheck
✓ CPUCheck
✓ DiskCheck
✓ MemoryCheck
✓ PortCheck
✓ Existing Kubernetes Cluster Check
✓ Completed Pre-Requisite Checks successfully
- prep-node
This command onboards a node. It installs platform9 packages on the host. After completion of this command, the node is available to be managed on the Platform9 control plane.
#pf9ctl prep-node --help
Prepare a node to be ready to be added to a Kubernetes cluster. Read more
at http://pf9.io/cli_clprep.
Usage:
pf9ctl prep-node [flags]
Flags:
-h, --help help for prep-node
-i, --ip strings IP address of host to be prepared
--mfa string MFA token
-p, --password string ssh password for the nodes (use 'single quotes' to pass password)
-c, --skip-checks Will skip optional checks if true
-s, --ssh-key string ssh key file for connecting to the nodes
-e, --sudo-pass string sudo password for user on remote host
-u, --user string ssh username for the nodes
Global Flags:
--log-dir string path to save logs
--no-prompt disable all user prompts
--verbose print verbose logs
prep-Node(Local)
#pf9ctl prep-node
✓ Loaded Config Successfully
✓ Removal of existing CLI
✓ Existing Platform9 Packages Check
✓ Required OS Packages Check
✓ SudoCheck
✓ CPUCheck
✓ DiskCheck
✓ MemoryCheck
✓ PortCheck
✓ Existing Kubernetes Cluster Check
✓ Completed Pre-Requisite Checks successfully
✓ Disabled swap and removed swap in fstab
✓ Platform9 packages installed successfully
✓ Initialised host successfully
✓ Host successfully attached to the Platform9 control-plane
prep-Node(Remote)
#pf9ctl prep-node -i 10.128.241.203 -u centos
✓ Loaded Config Successfully
You can choose either password or sshKey
Enter 1 for password and 2 for sshKey
Enter Option : 1
Enter password for remote host:
✓ Removal of existing CLI
✓ Existing Platform9 Packages Check
✓ Required OS Packages Check
✓ SudoCheck
✓ CPUCheck
✓ DiskCheck
✓ MemoryCheck
✓ PortCheck
✓ Existing Kubernetes Cluster Check
✓ Completed Pre-Requisite Checks successfully
✓ Disabled swap and removed swap in fstab
✓ Platform9 packages installed successfully
✓ Initialised host successfully
✓ Host successfully attached to the Platform9 control-plane
- bundle(SupportBundle)
This is used to gather and upload a support bundle(a folder containing logs for pf9 services and pf9ctl) to the S3 location.
#pf9ctl bundle --help
Gathers support bundle that includes logs for pf9 services and pf9ctl, uploads to S3
Usage:
pf9ctl bundle [flags]
Flags:
-h, --help help for bundle
-i, --ip strings IP address of host to be prepared
--mfa string MFA token
-p, --password string ssh password for the nodes (use 'single quotes' to pass password)
-s, --ssh-key string ssh key file for connecting to the nodes
-e, --sudo-pass string sudo password for user on remote host
-u, --user string ssh username for the nodes
Global Flags:
--log-dir string path to save logs
--no-prompt disable all user prompts
--verbose print verbose logs
bundle(Local)
#pf9ctl bundle
✓ Loaded Config Successfully
2021-05-17T06:45:03.95Z INFO ==========Uploading supportBundle to S3 bucket==========
✓ Succesfully uploaded pf9ctl supportBundle to loguploads.platform9.com bucket at https://s3-us-west-2.amazonaws.com/loguploads.platform9.com/pmkft-1614749234-62656.platform9.io/172.20.7.21/ location
bundle(Remote)
#pf9ctl bundle -i 10.128.241.203 -u centos
✓ Loaded Config Successfully
You can choose either password or sshKey
Enter 1 for password and 2 for sshKey
Enter Option : 1
Enter password for remote host:
2021-05-17T06:51:41.338Z INFO ==========Uploading supportBundle to S3 bucket==========
✓ Succesfully uploaded pf9ctl supportBundle to loguploads.platform9.com bucket at https://s3-us-west-2.amazonaws.com/loguploads.platform9.com/pmkft-1614749234-62656.platform9.io/172.20.7.104/ location
attach-node
#pf9ctl attach-node --help
Attach nodes to existing cluster. At a time, multiple workers but only one master can be attached
Usage:
pf9ctl attach-node [flags] cluster-name
Flags:
-h, --help help for attach-node
-m, --master-ip strings master node ip address
--mfa string MFA token
-u, --uuid string uuid of the cluster to attach the node to
-w, --worker-ip strings worker node ip address
Global Flags:
--log-dir string path to save logs
--no-prompt disable all user prompts
--verbose print verbose logs
#pf9ctl attach-node -m 172.20.7.66 -w 172.20.7.58 test-cluster
✓ Loaded Config Successfully
2021-05-26T11:58:01.9579Z INFO Worker node(s) [bf5364cf-e2fd-4500-97fb-0b01be26084f] attached to cluster
2021-05-26T11:58:03.6328Z INFO Master node(s) [615c1042-48a3-42e8-8003-ac135d12e6f4] attached to cluster
detach-node
#pf9ctl detach-node --help
Detach nodes from their clusters. If no ips are sent it will detach the node on which the command was run.
Usage:
pf9ctl detach-node [flags]
Flags:
-h, --help help for detach-node
--mfa string MFA token
-n, --node-ip strings node ip address
Global Flags:
--log-dir string path to save logs
--no-prompt disable all user prompts
--verbose print verbose logs
#pf9ctl detach-node
✓ Loaded Config Successfully
Starting detaching process
2021-11-08T08:35:14.1182Z INFO Node [9cfe32a2-6518-4b63-a55b-f9a9c1148e6a] detached from cluster
#pf9ctl detach-node -n ip
✓ Loaded Config Successfully
Starting detaching process
2021-11-08T08:35:14.1182Z INFO Node [9cfe32a2-6518-4b63-a55b-f9a9c1148e6a] detached from cluster
#pf9ctl detach-node -n ip1,ip2
✓ Loaded Config Successfully
Starting detaching process
2021-11-08T08:35:14.1182Z INFO Node [9cfe32a2-6518-4b63-a55b-f9a9c1148e6a] detached from cluster
2021-11-08T08:35:14.1182Z INFO Node [691a9feb-6b62-4235-bf27-208a14744843 detached from cluster
deauthorize-node
#pf9ctl deauthorize-node --help
Deauthorizes this node. It will warn the user if the node was a master node or a part of a single node cluster.
Usage:
pf9ctl deauthorize-node [flags]
Flags:
-h, --help help for deauthorize-node
-i, --ip string IP address of the host to be deauthorized
--mfa string MFA token
Global Flags:
--log-dir string path to save logs
--no-prompt disable all user prompts
--verbose print verbose logs
#pf9ctl deauthorize-node
✓ Loaded Config Successfully
Node deauthorization started....This may take a few minutes....Check the latest status in UI
authorize-node
#pf9ctl authorize-node --help
Authorizes this node
Usage:
pf9ctl authorize-node [flags]
Flags:
-h, --help help for authorize-node
-i, --ip string IP address of the host to be authorized
--mfa string MFA token
Global Flags:
--log-dir string path to save logs
--no-prompt disable all user prompts
--verbose print verbose logs
#pf9ctl authorize-node
✓ Loaded Config Successfully
Node authorization started....This may take a few minutes....Check the latest status in UI
delete-cluster
#pf9ctl delete-cluster --help
Deletes the cluster with the specified name. Additionally the user can pass the cluster UID instead of the name.
Usage:
pf9ctl delete-cluster [flags]
Flags:
-h, --help help for delete-cluster
--mfa string MFA token
-n, --name string clusters name
-i, --uuid string clusters uuid
Global Flags:
--no-prompt disable all user prompts
--verbose print verbose logs
#pf9ctl delete-cluster
2021-11-08T10:25:22.544Z FATAL You must pass a cluster name or the cluster uuid
#pf9ctl delete-cluster -n ClusterName
✓ Loaded Config Successfully
Cluster deletion started....This may take a few minutes.
#pf9ctl delete-cluster -i 023be0b0-1348-4d8a-a9b7-25bd4293cbbd
✓ Loaded Config Successfully
Cluster deletion started....This may take a few minutes.
decommission-node
#pf9ctl decommission-node --help
Removes the host agent package and decommissions this node from the Platform9 control plane.
Usage:
pf9ctl decommission-node [flags]
Flags:
-h, --help help for decommission-node
-i, --ip strings IP address of host to be decommissioned
--mfa string MFA token
-p, --password string ssh password for the nodes (use 'single quotes' to pass password)
-s, --ssh-key string ssh key file for connecting to the nodes
-u, --user string ssh username for the nodes
Global Flags:
--log-dir string path to save logs
--no-prompt disable all user prompts
--verbose print verbose logs
When node is connected to a cluster:
#pf9ctl decommission-node
✓ Loaded Config Successfully
Node is attached to test-2 cluster
2024-05-03T08:58:57.4328Z FATAL Node is still attached to a cluster. Please run detach-node command first and wait for the node to be completely removed from the cluster and only then run decommision-node command
When node is not connected to any cluster:
#pf9ctl decommission-node
✓ Loaded Config Successfully
Deauthorized node from UI
Removing pf9-hostagent (this might take a few minutes...)
Removed hostagent
Removing logs...
Running clean all
Removing pf9 HOME dir
Node decommissioning started....This may take a few minutes....Check the latest status in UI
check-amazon-provider
#pf9ctl check-amazon-provider -i iamUser -a access-key -s secret-key -r us-east-1
✓ ELB Access
✓ Route53 Access
✓ Availability Zones success
✓ EC2 Access
✓ VPC Access
✓ IAM Access
✓ Autoscaling Access
✓ EKS Access
check-google-provider
#pf9ctl check-google-provider -p /home/duser/Downloads/service-account.json -n testProject -e [email protected]
✓ Success roles/iam.serviceAccountUser
✓ Failed roles/container.admin
✓ Failed roles/compute.viewer
✓ Success roles/viewer
check-azure-provider
#pf9ctl check-google-provider -t tenantID -c clientID -s subscriptionID -k secretKey
✓ Has access
bootstrap
#pf9ctl bootstrap --help
Bootstrap a single node Kubernetes cluster with current node as the master node.
Usage:
pf9ctl bootstrap [flags] cluster-name
Examples:
pf9ctl bootstrap <clusterName> --pmk-version <version>
Required Flags:
--pmk-version string Kubernetes pmk version
Optional Flags:
--advanced-api-configuration string Allowed API groups and version. Option: default, all & custom
--allow-workloads-on-master Taint master nodes ( to enable workloads ), use either --allow-workloads-on-master or --allow-workloads-on-master=false to change (default true)
--api-server-flags strings Comma separated list of supported kube-apiserver flags, e.g: --request-timeout=2m0s,--kubelet-timeout=20s
--block-size string Block size determines how many Pods can run per node vs total number of nodes per cluster (default "26")
--container-runtime string The container runtime for the cluster (default "containerd")
--containers-cidr string CIDR for container overlay (default "10.20.0.0/16")
--controller-manager-flags strings Comma separated list of supported kube-controller-manager flags, e.g: --large-cluster-size-threshold=60,--concurrent-statefulset-syncs=10
--enable-kubeVirt Enables Kubernetes to run Virtual Machines within Pods. This feature is not recommended for production workloads, use either --enable-kubeVirt or --enable-kubeVirt=true to change
--enable-profile-engine Simplify cluster governance using the Platform9 Profile Engine, use either --enable-profile-engine or --enable-profile-engine=false to change (default true)
--etcd-backup Enable automated etcd backups on this cluster, use either --etcd-backup or --etcd-backup=false to change (default true)
--external-dns-name string External DNS for master VIP
-h, --help help for bootstrap
--http-proxy string Specify the HTTP proxy for this cluster. Format-> <scheme>://<username>:<password>@<host>:<port>, username and password are optional.
--interface-detction-method string Interface detection method for Calico CNI (default "first-found")
-i, --ip strings IP address of the host to be prepared
--ip-encapsulation string Encapsulates POD traffic in IP-in-IP between nodes (default "Always")
--master-virtual-interface string Physical interface for virtual IP association
--master-virtual-ip string Virtual IP address for cluster
--metallb-ip-range string Ip range for MetalLB
--mfa string MFA token
--monitoring Enable monitoring for this cluster, use either --monitoring or --monitoring=false to change (default true)
--mtu-size string Maximum Transmission Unit (MTU) for the interface (default "1440")
--nat int Packets destined outside the POD network will be SNAT'd using the node's IP (default 1)
--network-plugin string Specify network plugin ( Possible values: flannel or calico ) (default "calico")
--network-plugin-operator Will deploy Platform9 CRDs to enable multiple CNIs and features such as SR-IOV, use either --network-plugin-operator or --network-plugin-operator=true to change
--network-stack int 0 for ipv4 and 1 for ipv6
-p, --password string Ssh password for the node (use 'single quotes' to pass password)
--privileged Enable privileged mode for K8s API, use either --privileged or --privileged=false to change (default true)
-r, --remove-existing-pkgs Will remove previous installation if found, use either --remove-existing-pkgs or --remove-existing-pkgs=true to change
--reserved-cpu string Comma separated list of CPUs to be reserved for the system, e.g: 4-8,9-12
--scheduler-flags strings Comma separated list of supported Kube-scheduler flags, e.g: --kube-api-burst=120,--log_file_max_size=3000
--services-cidr string CIDR for services overlay (default "10.21.0.0/16")
-s, --ssh-key string Ssh key file for connecting to the node
-e, --sudo-pass string Sudo password for user on remote host
--tag string Add tag metadata to this cluster (key=value)
--topology-manager-policy string Topology manager policy (default "none")
--use-hostname Use node hostname for cluster creation, use either --use-hostname or --use-hostname=true to change
-u, --user string Ssh username for the node
Global Flags:
--log-dir string path to save logs
--no-prompt disable all user prompts
--verbose print verbose logs
#pf9ctl bootstrap testCluster --pmk-version 1.21.3-pmk.72
✓ Loaded Config Successfully
✓ Node is not onboarded and not attached to any cluster
✓ Removal of existing CLI
✓ Existing Platform9 Packages Check
✓ Required OS Packages Check
✓ SudoCheck
✓ CPUCheck
✓ DiskCheck
✓ MemoryCheck
✓ PortCheck
✓ Existing Kubernetes Cluster Check
✓ Check lock on dpkg
✓ Check lock on apt
✓ Check if system is booted with systemd
✓ Check time synchronization
✓ Check if firewalld service is not running
✓ Disabling swap and removing swap in fstab
✓ Completed Pre-Requisite Checks successfully
Prep local node as master node for kubernetes cluster (y/n): y
✓ Platform9 packages installed successfully
✓ Initialised host successfully
✓ Host successfully attached to the Platform9 control-plane
✓ Cluster creation started
✓ Host is connected
✓ Attached node to the cluster
✓ Bootstrap successfully finished
Cluster creation started....This may take a few minutes....Check the latest status in UI