Skip to content


A task is a group of actions that conforms a full backup of something (generates a file or a directory in the backup path for the current run). Tasks are defined inside a tasks definition file that contains a group of tasks that may share something in common - like all postgres backups will be in a file. The files can be defined using json or yaml syntax, and are located in the folder ${configPath}/tasks.

Tasks definition file

name: Good name that identifies this group of tasks (optional - file name will be used instead)
  one-variable: 1
  another-variable: true
    'yes': yes
    'no': no
inside: this/folder
  - name: Task 1
      more-variables: 'yes it is'
    stopOnFail: True
      - from-file: /etc/hosts
      - compress-gz: {}
      - to-file:
          path: hosts.gz
      ignore: false
  "name": "Good name that identifies this group of tasks (optional - file name will be used instead)",
  "env": {
    "one-variable": 1,
    "another-variable": true,
    "101-variable": {
      "yes": true,
      "no": false
  "inside": "this/folder",
  "tasks": [
      "name": "Task 1",
      "env": {
        "more-variables": "yes it is"
      "stopOnFail": false,
      "actions": [
        { "from-file": "/etc/hosts" },
        { "compress-gz": {} },
          "to-file": {
            "path": "hosts.gz"
      "cloud": {
        "ignore": false

One file will look like the above example. The example has all possible options that can have. Let's treat each of them.


The name of the tasks group/tasks definition file. To be able to identify each of them, a name is used. If name is not provided, then it will use the name of the file without extension as name.


If defined, then all files and directory created by output actions will be stored inside this path in the current backup path.


Defines variables that can be used in actions as parameters. They can also refer to secrets using #secret-name. If a variable matches with a key of a parameter for an action, this will be used as default value if the parameter is not defined in the action. These variables can be referenced inside a string by using ${VARIABLE_NAME}.


Defines all tasks that will be run in this file. One task contains its name and the actions to run. Optionally, it can define more variables in the env section. If stopOnFail is set to false and the task fails, it won't stop the whole backup. The cloud section is optional, and at the moment allows to ignore a task result to be uploaded.

The actions are defined with one item in the list by action, and to identify the action, the key of the dictionary is used:

- action-name: parameters
- action-name: parameters

Referring to secrets in env sections

Every time a variable has a string starting with #, those will be treated as secret references. When the task is going to run, the references are resolved with the real values of the secrets. The reference refers to a key path that can be found in a envDefs of any of the secret backends. Examples are better:

Imagine that the following secret backend config is set with this envDefs:

        user: 'secret/databases/postgres/pg-charizard-01#username'
        password: 'secret/databases/postgres/pg-charizard-01#password'
      encrypt-passphrase: 'secret/backups/vm-do-charizard-01/passphrase#passphrase'

So to refer to the user of postgres, this string will be used #postgres.user, as well as the password #postgres.password. For the passphrase, #encrypt-passphrase will be used.

This way, secrets are referred from the tasks using a key, and changing the path in the envDefs, will be changed in all the tasks that references the secret.

Environment variables in parameters

In the action parameters and the env sections, environment variables anv variables defined in env sections can be referred to customize even further the tasks. To reference a environment variable, just use it as ${ENV_VAR} in a string and it will be converted into its value. If a variable cannot be found, it will replace the token with an empty string (just delete the variable token). Variables in env sections can reference another variables in the same section because variable substitution is done just before running a task.

env sections and variable substitution

Strings and numbers are the only types supported for the variable substitution, any other type of variable will ignore it and a warning will be writen in the logs. When referencing variables from the same env section, order is important.


Simple file backup
  - name: File copy
      - from-file: /path/to/the/file
      - to-file:
          path: file
Run of a command, and the output is stored compressed in a file
  - name: Run a command
      - command:
        args: [command, parameter, parameter, parameter]
      - compress-xz:
          threads: 2
      - to-file:
          path: compressed-file.xz
Copy a folder
  - name: Copy a folder
      - from-directory: /path/to/the/folder
      - to-directory:
          path: folder
          reflink: yes
Copy a folder, archive it, compress it, encrypt it and stored it in a file
  - name: Do a lot of things with a folder
      - from-directory: /path/to/a/folder
      - tar: {}
      - compress-br: {}
      - encrypt-gpg:
          passphrase: '#gpg-password'
      - to-file:
Referrencing environment variables

```yaml tasks: - name: do stuff env: var1: yes var2: maybe ${var1} # This should be defined after var1 actions: - from-directory: '${HOME}/files' - tar: {} - compress-br: {} - to-file: path: ${var2}