Snowflake Object Lifecycle Orchestrator

Enterprise

Image	$DATAOPS_SNOWFLAKEOBJECTLIFECYCLE_RUNNER_IMAGE

The Snowflake Object Lifecycle Engine (SOLE) orchestrator is responsible for engaging with SOLE to process the Snowflake object configuration stored in the /dataops/snowflake project directory.

For more information, refer to the SOLE User Guide and the SOLE Reference Guide.

Usage

SOLE uses lifecycle actions to manage its processes and operations. As described in the SOLE Lifecycle Actions documentation, there are different lifecycle actions grouped into the following groups:

warning

PLAN-ALL and APPLY-ALL are coupled and apply only to non-cloned Snowflake databases.

Let's look at the code for each group as found in the DataOps Reference Project:

Setting up Snowflake with `AGGREGATE`

pipelines/includes/local_overrides/snowflake_lifecycle.yml
"Set Up Snowflake":
  extends:
    - .agent_tag
  stage: Snowflake Setup
  image: $DATAOPS_SNOWFLAKEOBJECTLIFECYCLE_RUNNER_IMAGE
  variables:
    LIFECYCLE_ACTION: AGGREGATE
    ARTIFACT_DIRECTORY: $CI_PROJECT_DIR/snowflake-artifacts
    CONFIGURATION_DIR: $CI_PROJECT_DIR/dataops/snowflake
  resource_group: $CI_JOB_NAME
  script:
    - /dataops
  artifacts:
    when: always
    paths:
      - $ARTIFACT_DIRECTORY
  icon: ${SNOWFLAKEOBJECTLIFECYCLE_ICON}
  rules:
    - if: '$DATAOPS_SOLE_ENABLE_PLAN_APPROVAL == "1" || $DATAOPS_SOLE_ENABLE_PLAN_APPROVAL == "true"'
      when: never
    - when: on_success

Cleaning up with `AGGREGATE-DESTROY`

pipelines/includes/local_overrides/snowflake_lifecycle.yml
"Tear Down Snowflake":
  extends:
    - .agent_tag
  stage: Clean Up
  image: $DATAOPS_SNOWFLAKEOBJECTLIFECYCLE_RUNNER_IMAGE
  variables:
    LIFECYCLE_ACTION: AGGREGATE-DESTROY
    ARTIFACT_DIRECTORY: $CI_PROJECT_DIR/snowflake-artifacts
    CONFIGURATION_DIR: $CI_PROJECT_DIR/dataops/snowflake
  resource_group: $CI_JOB_NAME
  script:
    - /dataops
  artifacts:
    when: always
    paths:
      - $ARTIFACT_DIRECTORY
  icon: ${SNOWFLAKEOBJECTLIFECYCLE_ICON}
  rules:
    ## Block this job from main (PROD) and qa (QA) branches
    - if: "$CI_COMMIT_REF_NAME == $DATAOPS_BRANCH_NAME_PROD || $CI_COMMIT_REF_NAME == $DATAOPS_BRANCH_NAME_QA"
      when: never
    ## For all other branches, enable this job to be run manually
    - when: manual

Setting up Snowflake with `PLAN-ALL`

pipelines/includes/local_overrides/snowflake_lifecycle.yml
"Plan Snowflake Setup":
  extends:
    - .agent_tag
  stage: Snowflake Setup
  image: $DATAOPS_SNOWFLAKELIFECYCLE_RUNNER_IMAGE
  variables:
    LIFECYCLE_ACTION: PLAN-ALL
    ARTIFACT_DIRECTORY: $CI_PROJECT_DIR/snowflake-artifacts
    CONFIGURATION_DIR: $CI_PROJECT_DIR/dataops/snowflake
  resource_group: $CI_JOB_NAME
  script:
    - /dataops
  artifacts:
    when: always
    paths:
      - $ARTIFACT_DIRECTORY
      - dataops/report/plan_viewer/
    name: "SOLE Plan Viewer"
    expose_as: "dataopsreport"
  icon: ${SNOWFLAKEOBJECTLIFECYCLE_ICON}
  rules:
    - if: '$DATAOPS_SOLE_ENABLE_PLAN_APPROVAL == "1" || $DATAOPS_SOLE_ENABLE_PLAN_APPROVAL == "true"'
      when: on_success
    - when: never

Applying actions with `APPLY-ALL`

pipelines/includes/local_overrides/snowflake_lifecycle.yml
"Apply Plan to Snowflake":
  extends:
    - .agent_tag
  stage: Snowflake Setup
  image: $DATAOPS_SNOWFLAKELIFECYCLE_RUNNER_IMAGE
  variables:
    LIFECYCLE_ACTION: APPLY-ALL
    ARTIFACT_DIRECTORY: $CI_PROJECT_DIR/snowflake-artifacts
    CONFIGURATION_DIR: $CI_PROJECT_DIR/dataops/snowflake
  resource_group: $CI_JOB_NAME
  needs:
    - job: Plan Snowflake Setup
  script:
    - /dataops
  artifacts:
    when: always
    paths:
      - $ARTIFACT_DIRECTORY
      - dataops/report/plan_viewer/
    name: "SOLE Plan Viewer"
    expose_as: "dataopsreport"
  icon: ${SNOWFLAKEOBJECTLIFECYCLE_ICON}
  rules:
    - if: '$DATAOPS_SOLE_ENABLE_PLAN_APPROVAL == "1" || $DATAOPS_SOLE_ENABLE_PLAN_APPROVAL == "true"'
      when: manual
    - when: never

Supported parameters

Parameter	Required/Default	Description
`CONFIGURATION_DIR`	REQUIRED. Defaults to `$CI_PROJECT_DIR/dataops/snowflake`	The project directory where the Snowflake configuration files are located.
`DATAOPS_SOLE_ACCOUNT`	REQUIRED—from connection parameters	The Snowflake account. If the account is a region-based account or the organization name is not used, the region must be added to this variable.
`DATAOPS_SOLE_USERNAME`	REQUIRED—for password-based authentication, from connection parameters	The Snowflake account username
`DATAOPS_SOLE_PASSWORD`	REQUIRED—for password-based authentication, from connection parameters	The Snowflake account password
`DATAOPS_SOLE_ROLE`	REQUIRED—from connection parameters	The Snowflake role that can run queries.
`DATAOPS_PREFIX`	REQUIRED—from project settings	This prefix is added to all account-level objects and databases (except the default database).
`DATAOPS_DATABASE`	REQUIRED—from project settings	The default database name - required if the default database is specified in the configuration as `{{env.DATAOPS_DATABASE}}`.
`DATAOPS_ENV_NAME`	REQUIRED—from project settings	This suffix is added to all account-level objects and databases (except the default database).
`LIFECYCLE_ACTION`	REQUIRED	See below for this parameter's valid values.
`ARTIFACT_DIRECTORY`	Optional. Defaults to `$CI_PROJECT_DIR/snowflake-provider`	The path to where the artifacts such as the resource files, import files, and state files are uploaded. It is essential to specify the `ARTIFACT_DIRECTORY` as an artifact in the related DataOps job.
`DISABLE_PERSISTENT_CACHE`	Optional	If set, it disables persistent cache usage by the orchestrator. The user must handle the generated file transfers between jobs.
`DATAOPS_DATABASE_MASTER`	Optional—from project settings	The `PROD` or main database - required if the production database is referenced in the configuration by the variable `{{env.DATAOPS_DATABASE_MASTER}}`.
`DATAOPS_ENV_NAME_PROD`	Optional. Defaults to `main`	Specifies the branch used as the production environment
`DATAOPS_NONDB_ENV_NAME`	Optional	The suffix override value - if specified, it overrides all branch-specific suffixes with its value.
`DATAOPS_PRESERVE_OBJECT_NAME`	Optional	If set, it enables special character support.
`DATAOPS_SOLE_DEBUG`	Optional	If set, it enables the debug logs for SOLE without exposing any credentials when writing to these logs.
`DATAOPS_SOLE_DISABLE_IMPLICIT_GRANTS`	Optional. Defaults to `false`	If set, it prevents SOLE from granting `USAGE` privilege on parent objects implicitly.
`DATAOPS_SOLE_ENABLE_PLAN_APPROVAL`	Optional	If set, it enables a coupled plan and apply process that allows you to review and approve modifications to a Snowflake database before applying them.
`DATAOPS_SOLE_MANAGE_REFERENCES`	Optional	If set, you can grant the `REFERENCES` privilege to the roles for materialized views.
`DATAOPS_SOLE_OVERRIDE`	Optional	If set, it overrides the object's delete/update prevention feature.
`DATAOPS_SOLE_WAREHOUSE`	Optional—from connection parameters	The Snowflake warehouse used for SOLE grant management.
`SNOWFLAKE_WAREHOUSE`	Optional—from connection parameters	The Snowflake warehouse used by SOLE to run hooks.
`DONT_USE_SOLE_GRANT_MANAGEMENT`	Optional	If set, it disables SOLE grant management for managing the privilege/role grants and revokes to use the Terraform Provider.
`LIFECYCLE_IMPORT_FILE`	Optional. Defaults to `resource_imports.json`	The file name containing the resource import statements.
`LIFECYCLE_MANAGE_OBJECT`	Optional	See below for this parameter's valid values. If no value is specified, all object groups are managed.
`LIFECYCLE_RESOURCE_FILE`	Optional. Defaults to `resources.tf.json`	The file name containing the resource definitions.
`LIFECYCLE_STATE_RESET`	Optional	If set, the local state is reset and all defined and managed objects are re-initialized.
`GENERATE_HOOKS_ARTIFACTS`	Optional	If set, it adds pre- and post-hooks to the generated report section of your pipeline. For more information about hooks, see Object Group Hooks.
`PROVIDER_DIR`	Optional. Defaults to `/snowflake-provider`	The provider configuration directory path that must contain a file named `provider.tf` with a list of Snowflake providers.
`SET_TERRAFORM_KEYS_TO_ENV`	Optional	If set, it exports SOLE credentials from the DataOps Vault to the pipeline environment.
`SOLE_LOG_CLEANUP`	Optional	If set, it removes all the log files from the host system's cache.
`SOLE_CACHE_CLEANUP`	Optional	If set, it removes the host system's cache.

`CONFIGURATION_DIR`

The CONFIGURATION_DIR parameter specifies your project directory where the Snowflake configuration files are located. Refer to the SOLE Managed Object Reference for supported configurations.

`LIFECYCLE_ACTION`

The valid values for this variable are as follows:

AGGREGATE - execute compile, validate, plan, and apply
COMPILE - only compile the Snowflake configuration found in the CONFIGURATION_DIR
VALIDATE - only validate the compiled resources
PLAN - only plan the validated configuration
APPLY - only apply the planned configuration
PLAN-ALL - execute compile, validate, and plan — but without "apply" contrary to AGGREGATE
APPLY-ALL - only apply the planned configuration by PLAN-ALL
AGGREGATE-DESTROY - execute compile, validate, plan-destroy, and destroy
PLAN-DESTROY - only plan how to tear down the Snowflake configuration
DESTROY - only execute the teardown plan

`LIFECYCLE_MANAGE_OBJECT`

The valid values for this variable are as follows:

ACCOUNT_LEVEL - only manage account-level objects
DATABASE - only manage databases
DATABASE_LEVEL - only manage database-level objects
GRANT - only manage grants

If no value is specified for this parameter, all object groups are managed in sequence.

`DATAOPS_SOLE_DEBUG`

Setting DATAOPS_SOLE_DEBUG to a value enables SOLE'S debug mode. This generates more log-output related to SOLE'S internal processes and is useful for debugging if a job fails or behaves unexpectedly. No credentials are logged when using DATAOPS_SOLE_DEBUG. The SOLE credentials are masked by a fixed 16-length character X.

Key-pair authentication

SOLE Orchestrator supports using key-pair authentication. To learn how to configure it, see key-pair authentication.

Example jobs

As described in the Usage section above, the SOLE orchestrator facilitates the implementation of the two aggregate lifecycle_actions, AGGREGATE and AGGREGATE-DESTROY. At the risk of duplicating content, providing context to the following example jobs is a good idea. Therefore, let's assume you need to create a new feature branch to develop a new report. Once the code has been written, it must be tested. To test this report, you need to run a DataOps pipeline, including the setup and tear-down Snowflake jobs.

Setting up Snowflake

The default from the DataOps Reference Project for setting up Snowflake uses the AGGREGATE lifecycle action.

pipelines/includes/local_overrides/snowflake_lifecycle.yml
"Set Up Snowflake":
  extends:
    - .agent_tag
  stage: "Snowflake Setup"
  image: $DATAOPS_SNOWFLAKEOBJECTLIFECYCLE_RUNNER_IMAGE
  variables:
    LIFECYCLE_ACTION: AGGREGATE
    ARTIFACT_DIRECTORY: $CI_PROJECT_DIR/snowflake-artifacts
    CONFIGURATION_DIR: $CI_PROJECT_DIR/dataops/snowflake
  resource_group: $CI_JOB_NAME
  script:
    - /dataops
  artifacts:
    when: always
    paths:
      - $ARTIFACT_DIRECTORY
  icon: ${SNOWFLAKEOBJECTLIFECYCLE_ICON}

Tearing down Snowflake

The default from the DataOps Reference Project for tearing down Snowflake for a feature branch uses the AGGREGATE-DESTROY lifecycle action.

pipelines/includes/local_overrides/snowflake_lifecycle.yml
"Tear Down Snowflake":
  extends:
    - .agent_tag
  stage: Clean Up
  image: $DATAOPS_SNOWFLAKEOBJECTLIFECYCLE_RUNNER_IMAGE
  variables:
    LIFECYCLE_ACTION: AGGREGATE-DESTROY
    ARTIFACT_DIRECTORY: $CI_PROJECT_DIR/snowflake-artifacts
    CONFIGURATION_DIR: $CI_PROJECT_DIR/dataops/snowflake
  resource_group: $CI_JOB_NAME
  script:
    - /dataops
  artifacts:
    when: always
    paths:
      - $ARTIFACT_DIRECTORY
  icon: ${SNOWFLAKEOBJECTLIFECYCLE_ICON}
  rules:
    ## Block this job from main (PROD) and qa (QA) branches
    - if: "$CI_COMMIT_REF_NAME == $DATAOPS_BRANCH_NAME_PROD || $CI_COMMIT_REF_NAME == $DATAOPS_BRANCH_NAME_QA"
      when: never
    ## For all other branches, enable this job to be run manually
    - when: manual

Individual jobs

For further examples on how to use the lifecycle actions and the object hierarchies with the SOLE orchestrator, refer to the lifecycle action jobs user guide.

Enterprise

Usage​

Setting up Snowflake with AGGREGATE​

Cleaning up with AGGREGATE-DESTROY​

Setting up Snowflake with PLAN-ALL​

Applying actions with APPLY-ALL​

Supported parameters​

CONFIGURATION_DIR​

LIFECYCLE_ACTION​

LIFECYCLE_MANAGE_OBJECT​

DATAOPS_SOLE_DEBUG​

Key-pair authentication​

Example jobs​

Setting up Snowflake​

Tearing down Snowflake​

Individual jobs​