Files
toallab-automation/roles/ansible-network.network-engine/docs/directives/template_directives.md
Patrick Toal 6e2205a046 Adding Netbox
2019-05-06 00:34:45 -04:00

8.8 KiB

CLI Template Directives

Note network_template lookup plugin is deprecated in v2.7.3 and will be removed in version v2.7.7 i.e, four releases from the deprecation version.

The network_template module supports a number of keyword based objectives that handle how to process the template. Templates are broken up into a series of blocks that process lines. Blocks are logical groups that have a common set of properties in common.

Blocks can also include other template files and are processed in the same manner as lines. See includes below for a description on how to use the include directive.

The template module works by processing the lines directives in sequential order. The module will attempt to template each line in the lines directive and, if successful, add the line to the final output. Values used for variable substitution come from the host facts. If the line could not be successfully templated, the line is skipped and a warning message is displayed that the line could not be templated.

There are additional directives that can be combined to support looping over lists and hashes as well as applying conditional statements to blocks, lines and includes.

name

Entries in the template may contain a name field. The name field is used to provide a description of the entry. It is also used to provide feedback when processing the template to indicate when an entry is skipped or fails.

lines

The lines directive provides an ordered list of statements to attempt to template. Each entry in the lines directive will be evaluated for variable substitution. If the entry can be successfully templated, then the output will be added to the final set of entries. If the entry cannot be successfully templated, then the entry is ignored (skipped) and a warning message is provided. If the entry in the lines directive contains only static text (no variables), then the line will always be processed.

The lines directive also supports standard Jinja2 filters as well as any Ansible specific Jinja2 filters. For example, lets assume we want to add a default value if a more specific value was not assigned by a fact.

- name: render the system hostname
  lines:
    - "hostname {{ hostname | default(inventory_hostname_short }}"

block

A group of lines directives can be combined into a block directive. These block directives are used to apply a common set of values to one or more lines or includes entries.

For instance, a block directive that contains one or more lines entries could be use the same set of loop values or have a common when conditional statement applied to them.

include

Sometimes it is advantageous to break up templates into separate files and combine them. The include directive will instruct the current template to load another template file and process it.

The include directive also supports variable substitution for the provided file name and can be processed with the loop and when directives.

when

The when directive allows for conditional statements to be applied to a set of lines, a block and/or the include directive. The when statement is evaluated prior to processing the statements and if the condition is true, the statements will attempt to be templated. If the statement is false, the statements are skipped and a message returned.

loop

Depending on the input facts, sometimes it is necessary to iterate over a set of statements. The loop directive allows the same set of statements to be processed in such a manner. The loop directive takes, as input, the name of a fact that is either a list or a hash and iterates over the statements for each entry.

When the provided fact is a list of items, the value will be assigned to a variable called item and can be referenced by the statements.

When the provided fact is a hash of items, the hash key will be assigned to the item.key variable and the hash value will be assigned to the item.value variable. Both can then be referenced by the statements.

loop_control

The loop_control directive allows the template to configure aspects related to how loops are process. This directive provides a set of suboptions to configure how loops are processed.

loop_var

The loop_var directive allows the template to override the default variable name item. This is useful when handling nested loops such that both inner and outer loops values can be accessed.

When setting the loop_var to some string, the string will replace item as the variable name used to access the values.

For example, lets assume instead of using item, we want to use a different variable name such as entry:

- name: render entries
  lines:
    - "hostname {{ entry.hostname }}"
    - "domain-name {{ entry.domain_name }}"
  loop: "{{ system }}"
  loop_control:
    loop_var: entry

join

When building template statements that include optional values for configuration, the join directive can be useful. The join directive instructs the template to combine the templated lines together into a single string to insert into the configuration.

For example, lets assume there is a need to add the following statement to the configuration:

ip domain-name ansible.com vrf management
ip domain-name redhat.com

To support templating the above lines, the facts might include the domain-name and the vrf name values. Here is the example facts:

---
system:
  - domain_name: ansible.com
    vrf: management
  - domain_name redhat.com

And the template statement would be the following:

- name: render domain-name
  lines:
    - "ip domain-name {{ item.domain_name }}"
    - "vrf {{ item.vrf }}"
  loop: "{{ system }}"
  join: yes

When this entry is processed, the first iteration will successfully template both lines and add ip domain-name ansible.com vrf management to the output.

When the second entry is processed, the first line will be successfully templated but since there is no management key, the second line will return a null value. The final line added to the configuration will be ip domain-name redhat.com.

If the join directive had been omitted, then the final set of configuration statements would be as follows:

ip domain-name ansible.com
vrf management
ip domain-name redhat.com

join_delimiter

When the join delimiter is used, the templated values are combined into a single string that is added to the final output. The lines are joined using a space. The delimiting character used when processing the join can be modified using join_delimiter directive.

Here is an example of using the this directive. Take the following entry:

- name: render domain-name
  lines:
    - "ip domain-name {{ item.domain_name }}"
    - "vrf {{ item.vrf }}"
  loop: "{{ system }}"
  join: yes
  join_delimiter: ,

When the preceding statements are processed, the final output would be (assuming all variables are provided):

ip domain-name ansible.com,vrf management
ip domain-name redhat.com

indent

The indent directive is used to add one or more leading spaces to the final templated statement. It can be used to recreated a structured configuration file.

Take the following template entry as an example:

- name: render the interface context
  lines: "interface Ethernet0/1"

- name: render the interface configuration
  lines:
    - "ip address 192.168.10.1/24"
    - "no shutdown"
    - "description this is an example"
  indent: 3

- name: render the interface context
  lines: "!"

Then the statements above are processed, the output will look like the following:

interface Ethernet0/1
   ip address 192.168.10.1/24
   no shutdown
   description this is an example
!

required

The required directive specifies that all of the statements must be templated otherwise a failure is generated. Essentially it is a way to make certain that the variables are defined.

For example, take the following:

- name: render router ospf context
  lines:
    - "router ospf {{ process_id }}"
  required: yes

When the above is processed, if the variable process_id is not present, then the statement cannot be templated. Since the required directive is set to true, the statement will cause the template to generate a failure message.

missing_key

By default, when statements are processed and a variable is undefined, the module will simply display a warning message to the screen. In some cases, it is desired to either suppress warning messages on a missing key or to force the module to fail on a missing key.

To change the default behaviour, use the missing_key directive. This directive accepts one of three choices:

  • ignore
  • warn (default)
  • fail

The value of this directive will instruct the template how to handle any condition where the desired variable is undefined.