Nauta uses Ansible* for certain provisioning tasks during installation. You must create (or modify) an Ansible inventory file to match your hardware configuration. Nauta will look for your inventory file at the location defined in the ENV_INVENTORY environment variable (see Installation Process for more information).
Your Nauta cluster will contain one Master node and one or more Worker nodes. Each of these nodes must be specified in the Inventory file.
Below is an example of Inventory file and shows one Master Node and five Worker nodes. Your configuration may differ from the example shown.
Note: Ansible uses the YAML* format.
**[master]**
master-0 ansible_ssh_host=192.168.100 ansible_ssh_user=root ansible_ssh_pass=YourPassword internal_interface=em2 external_interface=em3 local_data_device=/dev/sdb1
**[worker]**
worker-0 ansible_ssh_host=192.168.100.61 ansible_ssh_user=root ansible_ssh_pass=YourPassword internal_interface=p3p1 external_interface=em1
worker-1 ansible_ssh_host=192.168.100.55 ansible_ssh_user=root ansible_ssh_pass=YourPassword internal_interface=p3p1 external_interface=em1
worker-3 ansible_ssh_host=192.168.100.106 ansible_ssh_user=root ansible_ssh_ pass=YourPassword internal_interface=p3p1 external_interface=em1
worker-4 ansible_ssh_host=192.168.100.107 ansible_ssh_user=root ansible_ssh_ pass=YourPassword internal_interface=p3p1 external_interface=em1The file contains two sections:
**[master]**Contains a description of a master node. This section must contain exactly one row.**[worker]**Contains descriptions of workers. Each worker is described in one row. In this section, it can have one or many rows depending on a structure of a cluster.
Each row describes a server (playing either the role of Master or Worker depending on which section the row is in). For each server, the inventory file must define a series of values that tells Nauta where to find the server, how to log into it, and so on. The format for each row is as follows:
[SERVER_NAME] [VAR_NAME1]=[VAR_VALUE1] [VAR_NAME2]=[VAR_VALUE2] [VAR_NAME3]=[VAR_VALUE3]...
NOTE: SERVER_NAME must conform to standard host naming rules - each element of the hostname must be from 1 to
63 characters long and the entire hostname, including the dots, can be at most 253 characters long. Valid characters
for hostnames are ASCII(7) letters from a to z, the digits from 0 to 9, and the hyphen. A hostname may not start with
a hyphen.
The table below lists all the variables understood by Nauta's inventory system. Some variables are required for all servers in the inventory, some are only required for some, and some variables are entirely optional.
| Variable Name | Description | Req? | Type | Default | Used When | Value |
|---|---|---|---|---|---|---|
| ansible_ssh_user | The user name must be the same for master and worker nodes. Note: If an Administrator decides to choose something other than root for Ansible SSH user, the user must be configured in sudoers file with NOPASSWD option. Refer to Ansible Inventory documentation for more information. | Yes | string | none | always | username |
| ansible_ssh_pass | The SSH password to use. | Yes | string | none | always | Password |
| ansible_ssh_host | The name (DNS name) or IP Address of the host to connect to. | Yes | IPaddr | none | always | IP Address |
| ansible_ssh_port | The SSH port number, if not defined 22 is used. | No | int | 22 | not using 22 | Port Address |
| ansible_ssh_private_key_file | This is a Private Key file used by SSH. | No | string | none | using a keyfile | filename |
| internal_interface | This is used for internal communication between Kubernetes* processes and pods. All interfaces (both external and internal) are Ethernet interfaces. | Yes | string | none | always for both for master and worker nodes | Interface name |
| local_data_device | This device is used for Nauta internal data and NFS data in case of local NFS. | Yes | string | none | used with master nodes | Path to block device |
| local_device | This device is used for Nauta internal data and NFS data in case of local NFS. | Yes | string | none | used with master nodes | Path to block device |
| local_data_path | This is used as the mountpoint for local_data_device |
No | string | none | used with master nodes | Absolute path where data is located in file system |
| external_interface | This is used for external network communication. | Yes | string | none | always for both for master and worker nodes | Interface name |