Nornir Service Test Task

The Nornir Service test task designed to facilitate the execution of network tests. This task provides network operations engineers and network automation developers with tools to validate network configurations, ensure compliance, and monitor network performance. By leveraging the capabilities of the Nornir service, users can automate testing process, identify issues proactively, and maintain a robust network infrastructure.

Nornir service test task uses Nornir TestsProcessor to run the tests and support test suites definition in YAML format, where test suite YAML files can be stored on and sourced from broker.

Nornir Test Sample Usage

Nornir service test task uses suites in YAML format to define tests, sample tests suite:

suite_3.txt

- name: Check ceos version
  task: "show version"
  test: contains
  pattern: "4.30.0F"
- name: Check NTP status
  test: ncontains
  pattern: "unsynchronised"
  task: "show ntp status"
- name: Check Mgmt Interface Status
  test: contains
  pattern: "is up, line protocol is up"
  task: "show interface management0" 

File suite_3.txt stored on broker and downloaded by Nornir service prior to running tests, below is an example of how to run the tests suite.

Example

CLIPython

C:\nf>nfcli
Welcome to NorFab Interactive Shell.
nf#
nf#nornir
nf[nornir-test]#
nf[nornir-test]#suite nf://nornir_test_suites/suite_3.txt FC spine,leaf
--------------------------------------------- Job Events -----------------------------------------------
07-Jan-2025 18:44:35 0c3309c54ee44397b055257a0d442e62 job started
07-Jan-2025 18:44:35.207 nornir nornir-worker-1 ceos-spine-1, ceos-spine-2 task started - 'netmiko_send_commands'
07-Jan-2025 18:44:35.211 nornir nornir-worker-2 ceos-leaf-1, ceos-leaf-2, ceos-leaf-3 task started - 'netmiko_send_commands'
<omitted for brevity>
07-Jan-2025 18:44:36 0c3309c54ee44397b055257a0d442e62 job completed in 1.391 seconds

--------------------------------------------- Job Results --------------------------------------------

+----+--------------+-----------------------------+----------+-------------------+
|    | host         | name                        | result   | exception         |
+====+==============+=============================+==========+===================+
|  0 | ceos-leaf-1  | Check ceos version          | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
|  1 | ceos-leaf-1  | Check NTP status            | FAIL     | Pattern in output |
+----+--------------+-----------------------------+----------+-------------------+
|  2 | ceos-leaf-1  | Check Mgmt Interface Status | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
|  3 | ceos-leaf-2  | Check ceos version          | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
|  4 | ceos-leaf-2  | Check NTP status            | FAIL     | Pattern in output |
+----+--------------+-----------------------------+----------+-------------------+
|  5 | ceos-leaf-2  | Check Mgmt Interface Status | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
|  6 | ceos-leaf-3  | Check ceos version          | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
|  7 | ceos-leaf-3  | Check NTP status            | FAIL     | Pattern in output |
+----+--------------+-----------------------------+----------+-------------------+
|  8 | ceos-leaf-3  | Check Mgmt Interface Status | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
|  9 | ceos-spine-1 | Check ceos version          | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
| 10 | ceos-spine-1 | Check NTP status            | FAIL     | Pattern in output |
+----+--------------+-----------------------------+----------+-------------------+
| 12 | ceos-spine-2 | Check ceos version          | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
| 12 | ceos-spine-2 | Check ceos version          | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
| 12 | ceos-spine-2 | Check ceos version          | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
| 12 | ceos-spine-2 | Check ceos version          | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
| 12 | ceos-spine-2 | Check ceos version          | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
| 13 | ceos-spine-2 | Check NTP status            | FAIL     | Pattern in output |
+----+--------------+-----------------------------+----------+-------------------+
| 12 | ceos-spine-2 | Check ceos version          | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
| 12 | ceos-spine-2 | Check ceos version          | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
| 13 | ceos-spine-2 | Check NTP status            | FAIL     | Pattern in output |
+----+--------------+-----------------------------+----------+-------------------+
| 14 | ceos-spine-2 | Check Mgmt Interface Status | PASS     |                   |
+----+--------------+-----------------------------+----------+-------------------+
nf[nornir-test]#
nf[nornir-test]#top
nf#

Demo

In this example:

• nfcli command starts the NorFab Interactive Shell.
• nornir command switches to the Nornir sub-shell.
• test command switches to the test task sub-shell.
• suite argument refers to a path for suite_3.txt file with a set of tests to run.
• Devices filtered using FC - "Filter Contains" Nornir hosts targeting filter to only run tests on devices that contain spine or leaf in their hostname.

inventory.yaml should be located in same folder where we start nfcli, unless nfcli -i path_to_inventory.yaml flag used. Refer to Getting Started section on how to construct inventory.yaml file

This code is complete and can run as is

import pprint

from norfab.core.nfapi import NorFab

if __name__ == '__main__':
    nf = NorFab(inventory="inventory.yaml")
    nf.start()

    client = nf.make_client()

    res = client.run_job(
        service="nornir",
        task="test",
        kwargs={
            "suite": "nf://nornir_test_suites/suite_3.txt",
            "FC": "spine,leaf"          
        }
    )

    pprint.pprint(res)

    nf.destroy()

Refer to Getting Started section on how to construct inventory.yaml file.

Formatting Tests Output

NorFab interactive shell allows you to format the results of network tests into text tables. This is particularly useful for presenting test results in a clear and organized manner, making it easier to analyze and interpret the data. The NorFab interactive shell supports the table command, which relies on the tabulate module to generate text tables. By outputting test results in table format, you can quickly identify issues and take appropriate action.

Using Jinja2 Templates to Generate Tests

Using Jinja2 Templates enables you to create dynamic test suites based on variables defined in your inventory or passed as job data. This approach allows you to tailor tests to specific devices or scenarios, ensuring that the tests are relevant and accurate. Jinja2 templates provide a powerful way to automate the creation of complex test cases, incorporating conditional logic, loops, and other advanced features to meet your testing requirements.

Templating Tests with Inline Job Data

Inline Job Data allows you to define test parameters directly within the job_data argument, making it easy to customize tests on the fly. This feature is particularly useful for scenarios where test parameters need to be adjusted frequently or based on specific conditions. By templating tests with inline job data, you can ensure that your tests are always up-to-date and aligned with the current network state.

Using Dry Run

The Using Dry Run feature allows you to generate the content of network test suites without actually performing any actions on the devices. This is useful for validation purposes, as it enables you to verify the correctness of your tests before running them. By using dry run, you can identify potential issues and make necessary adjustments, ensuring that your tests will execute successfully when run for real.

Running a Subset of Tests

Running a Subset of Tests allows you to execute only a specific set of tests, rather than running the entire test suite. This is useful for targeted testing, such as validating changes in a particular part of the network configuration or focusing on specific devices features. By running a subset of tests, you can save time and resources, while still ensuring that critical aspects of the network are thoroughly tested.

Returning Only Failed Tests

Returning only failed tests enables you to filter the test results to show only the tests that have failed. This is particularly useful for quickly identifying and addressing issues, as it allows you to focus on the areas that require attention. By returning only failed tests, you can streamline the troubleshooting process and ensure that network problems are resolved efficiently.

NORFAB Nornir Test Shell Reference

The NORFAB Nornir Test Shell Reference provides a comprehensive set of command options for the Nornir test task. These commands allow you to control various aspects of the test execution, such as setting job timeouts, filtering devices, adding task details to results, and configuring retry mechanisms. By leveraging these command options, you can tailor the behavior of the tests to meet your specific network management needs, ensuring that your network remains reliable and performant.

NorFab shell supports these command options for Nornir test task:

nf#man tree nornir.test
root
└── nornir:    Nornir service
    └── test:    Run network tests
        ├── timeout:    Job timeout
        ├── workers:    Filter worker to target, default 'all'
        ├── add_details:    Add task details to results
        ├── run_num_workers:    RetryRunner number of threads for tasks execution
        ├── run_num_connectors:    RetryRunner number of threads for device connections
        ├── run_connect_retry:    RetryRunner number of connection attempts
        ├── run_task_retry:    RetryRunner number of attempts to run task
        ├── run_reconnect_on_fail:    RetryRunner perform reconnect to host on task failure
        ├── run_connect_check:    RetryRunner test TCP connection before opening actual connection
        ├── run_connect_timeout:    RetryRunner timeout in seconds to wait for test TCP connection to establish
        ├── run_creds_retry:    RetryRunner list of connection credentials and parameters to retry
        ├── tf:    File group name to save task results to on worker file system
        ├── tf_skip_failed:    Save results to file for failed tasks
        ├── diff:    File group name to run the diff for
        ├── diff_last:    File version number to diff, default is 1 (last)
        ├── progress:    Emit execution progress
        ├── headers:    Table headers
        ├── headers_exclude:    Table headers to exclude
        ├── sortby:    Table header column to sort by
        ├── reverse:    Table reverse the sort by order
        ├── FO:    Filter hosts using Filter Object
        ├── FB:    Filter hosts by name using Glob Patterns
        ├── FH:    Filter hosts by hostname
        ├── FC:    Filter hosts containment of pattern in name
        ├── FR:    Filter hosts by name using Regular Expressions
        ├── FG:    Filter hosts by group
        ├── FP:    Filter hosts by hostname using IP Prefix
        ├── FL:    Filter hosts by names list
        ├── FM:    Filter hosts by platform
        ├── FX:    Filter hosts excluding them by name
        ├── FN:    Negate the match
        ├── hosts:    Filter hosts to target
        ├── *suite:    Nornir suite nf://path/to/file.py, default 'PydanticUndefined'
        ├── dry_run:    Return produced per-host tests suite content without running tests
        ├── subset:    Filter tests by name
        ├── failed_only:    Return test results for failed tests only
        ├── remove_tasks:    Include/Exclude tested task results
        └── job_data:    Path to YAML file with job data
nf#

* - mandatory/required command argument

Python API Reference

Function to tests data obtained from devices.

Parameters:

Name	Type	Description	Default
suite	Union[list, str]	path to YAML file with tests	required
dry_run	bool	if True, returns produced per-host tests suite content only	False
subset	str	list or string with comma separated non case sensitive glob patterns to filter tests' by name, subset argument ignored by dry run	None
failed_only	bool	if True returns test results for failed tests only	False
remove_tasks	bool	if False results will include other tasks output	True
return_tests_suite	bool	if True returns rendered per-host tests suite content in addition to test results using dictionary with results and suite keys	False
job_data	str	URL to YAML file with data or dictionary/list of data to pass on to Jinja2 rendering context	None
kwargs		any additional arguments to pass on to Nornir service task	{}

Source code in norfab\workers\nornir_worker.py

def test(
    self,
    suite: Union[list, str],
    subset: str = None,
    dry_run: bool = False,
    remove_tasks: bool = True,
    failed_only: bool = False,
    return_tests_suite: bool = False,
    job_data: str = None,
    **kwargs,
) -> dict:
    """
    Function to tests data obtained from devices.

    :param suite: path to YAML file with tests
    :param dry_run: if True, returns produced per-host tests suite content only
    :param subset: list or string with comma separated non case sensitive glob
        patterns to filter tests' by name, subset argument ignored by dry run
    :param failed_only: if True returns test results for failed tests only
    :param remove_tasks: if False results will include other tasks output
    :param return_tests_suite: if True returns rendered per-host tests suite
        content in addition to test results using dictionary with ``results``
        and ``suite`` keys
    :param job_data: URL to YAML file with data or dictionary/list of data
        to pass on to Jinja2 rendering context
    :param kwargs: any additional arguments to pass on to Nornir service task
    """
    tests = {}  # dictionary to hold per-host test suites
    add_details = kwargs.get("add_details", False)  # ResultSerializer
    to_dict = kwargs.get("to_dict", True)  # ResultSerializer
    filters = {k: kwargs.pop(k) for k in list(kwargs.keys()) if k in FFun_functions}
    ret = Result(task=f"{self.name}:test", result={} if to_dict else [])
    suites = {}  # dictionary to hold combined test suites

    self.nr.data.reset_failed_hosts()  # reset failed hosts
    filtered_nornir = FFun(self.nr, **filters)  # filter hosts

    # check if no hosts matched
    if not filtered_nornir.inventory.hosts:
        msg = (
            f"{self.name} - nothing to do, no hosts matched by filters '{filters}'"
        )
        log.debug(msg)
        ret.messages.append(msg)
        if return_tests_suite is True:
            ret.result = {"test_results": [], "suite": {}}
        return ret

    # download job data
    job_data = self.load_job_data(job_data)

    # generate per-host test suites
    for host_name, host in filtered_nornir.inventory.hosts.items():
        # render suite using Jinja2
        try:
            rendered_suite = self.render_jinja2_templates(
                templates=[suite],
                context={
                    "host": host,
                    "norfab": self.client,
                    "nornir": self,
                    "job_data": job_data,
                },
                filters=self.add_jinja2_filters(),
            )
        except Exception as e:
            msg = f"{self.name} - '{suite}' Jinja2 rendering failed: '{e}'"
            raise RuntimeError(msg)
        # load suit using YAML
        try:
            tests[host_name] = yaml.safe_load(rendered_suite)
        except Exception as e:
            msg = f"{self.name} - '{suite}' YAML load failed: '{e}'"
            raise RuntimeError(msg)

    # validate tests suite
    try:
        _ = modelTestsProcessorSuite(tests=tests)
    except Exception as e:
        msg = f"{self.name} - '{suite}' suite validation failed: '{e}'"
        raise RuntimeError(msg)

    # download pattern, schema and custom function files
    for host_name in tests.keys():
        for index, item in enumerate(tests[host_name]):
            for k in ["pattern", "schema", "function_file"]:
                if self.is_url(item.get(k)):
                    item[k] = self.fetch_file(
                        item[k], raise_on_fail=True, read=True
                    )
                    if k == "function_file":
                        item["function_text"] = item.pop(k)
            tests[host_name][index] = item

    # save per-host tests suite content before mutating it
    if return_tests_suite is True:
        return_suite = copy.deepcopy(tests)

    log.debug(f"{self.name} - running test '{suite}', is dry run - '{dry_run}'")
    # run dry run task
    if dry_run is True:
        result = filtered_nornir.run(
            task=nr_test, name="tests_dry_run", ret_data_per_host=tests
        )
        ret.result = ResultSerializer(
            result, to_dict=to_dict, add_details=add_details
        )
    # combine per-host tests in suites based on task and arguments
    # Why - to run tests using any nornir service tasks with various arguments
    else:
        for host_name, host_tests in tests.items():
            for test in host_tests:
                dhash = hashlib.md5()
                test_args = test.pop("norfab", {})
                nrtask = test_args.get("nrtask", "cli")
                assert nrtask in [
                    "cli",
                    "network",
                    "cfg",
                    "task",
                ], f"{self.name} - unsupported NorFab Nornir Service task '{nrtask}'"
                test_json = json.dumps(test_args, sort_keys=True).encode()
                dhash.update(test_json)
                test_hash = dhash.hexdigest()
                suites.setdefault(test_hash, {"params": test_args, "tests": {}})
                suites[test_hash]["tests"].setdefault(host_name, [])
                suites[test_hash]["tests"][host_name].append(test)
        log.debug(
            f"{self.name} - combined per-host tests suites based on NorFab Nornir Service task and arguments:\n{suites}"
        )
        # run test suites collecting output from devices
        for tests_suite in suites.values():
            nrtask = tests_suite["params"].pop("nrtask", "cli")
            function_kwargs = {
                **tests_suite["params"],
                **kwargs,
                **filters,
                "tests": tests_suite["tests"],
                "remove_tasks": remove_tasks,
                "failed_only": failed_only,
                "subset": subset,
            }
            result = getattr(self, nrtask)(
                **function_kwargs
            )  # returns Result object
            # save test results into overall results
            if to_dict == True:
                for host_name, host_res in result.result.items():
                    ret.result.setdefault(host_name, {})
                    ret.result[host_name].update(host_res)
            else:
                ret.result.extend(result.result)

    # check if need to return tests suite content
    if return_tests_suite is True:
        ret.result = {"test_results": ret.result, "suite": return_suite}

    return ret