eos_validate_state - Preview Integration with ANTA¶
Warning
eos_validate_state integration with ANTA is in preview. Everything is subject to change. If you have any questions, please leverage the GitHub discussions board
Warning
ANTA version has been bumped to 0.14.0. Please make sure you are running this exact version! For more details, please see the installation section.
Overview¶
eos_validate_state is a role leveraged to validate Arista EOS devices’ operational states.
eos_validate_state role:
- Consumes structured EOS configuration file, the same input as the role eos_cli_config_gen. This input is considered the system of record (the desired state).
- Connects to EOS devices to collect operational states (actual state). This requires access to the configured devices.
- Compares the actual states against the desired state.
- Generates CSV and Markdown reports of the results.
Known limitations¶
- Loose mode to ignore playbook errors is no longer supported in ANTA mode.
- ANTA mode exclusively supports the newer “list-of-dicts” data models in the structured configuration file input. For further details, consult the AVD 4.x.x porting guides.
Roadmap¶
Note
Subject to change. No commitments implied.
- Add more tests generated from the structured configuration
Tip
You can now provide your own custom ANTA catalogs to the AVD eos_validate_state
role! Please refer to the Custom ANTA catalog section for more details.
Expected changes¶
- You should expect faster execution, and if not please report on the GitHub discussions board
- Hardware tests are now collapsed.
- Some description of tests have been updated to be more precise.
- Sorting of the test results is now done per device as opposed to per category.
- Tests skipped by ANTA will be marked as
SKIPPED
in the final reports. - All tests will be removed from the catalog for a device flagged as undeployed using the host level variable
is_deployed: false
. Additionally, all tests take into account theis_deployed
variable value and remove tests accordingly.
Warning
Tests can also be automatically removed from the catalogs depending on the structured configuration of the devices. ANTA is therefore not aware of these tests and they will not appear in the final report. For example, the AvdTestMLAG
tests will not be present in the test catalog of a device that does not have an MLAG configuration in its structured configuration.
- BGP tests will only run if
service_routing_protocols_model
is set tomulti-agent
in the structured configuration file.
Note
Starting from version 4.30.1F, service_routing_protocols_model
is preset to multi-agent
by default on EOS devices.
- Inband management reachability test has been refactored to support the new AVD 4.x.x inband management data model.
How to run eos_validate_state in ANTA mode¶
-
Install the “anta” Python package (this is not part of the
requirements.txt
):
-
Run eos_validate_state playbook by setting the variable
use_anta=true
.This can be set for instance in your group_vars or under the task in your playbook.
If you
use_anta=false
which is the default, the current version ofeos_validate_state
leveraging Ansible asserts will be run.
-
Ansible tags are supported for backwards compatibility until AVD version 5.0.0. To run/skip tests use
--tags
or--skip-tags
.
-
You can now run the eos_validate_state role in check_mode. This will produce a report of tests that will be performed without running the tests on your network. Tests will be marked as
NOT RUN
in the final reports.
-
You can increase the Ansible verbosity by adding multiple
-v
when running the playbook. This will give you visibility on which test categories are being removed from a device’s catalog by AVD according to the structured configurations.
Info
ANTA mode also supports other functionalities. For more details, please refer to the input variables below.
Test Categories¶
Note
New tests are marked with the (New) string.
- AvdTestHardware (Ansible tags:
hardware
,platform_information
)- VerifyEnvironmentPower: Validate environment power supplies status.
- VerifyEnvironmentCooling: Validate environment fan status.
- VerifyTemperature: Validate environment temperature.
- VerifyTransceiversManufacturers: Validate transceivers manufacturer.
- AvdTestNTP (Ansible tags:
ntp
)- VerifyNTP: Validate NTP status.
- AvdTestInterfacesState (Ansible tags:
interfaces_state
)- VerifyInterfacesStatus: Validate interfaces status.
- Ethernet interfaces
- Port-channel interfaces
- Vlan interfaces
- Loopback interfaces
- Vxlan1 interface
- DPS interfaces
- VerifyInterfacesStatus: Validate interfaces status.
- AvdTestP2PIPReachability (Ansible tags:
ip_reachability
)- VerifyReachability: Validate IP reachability for point-to-point l3 ethernet interfaces.
- AvdTestInbandReachability (Ansible tags:
loopback_reachability
,loopback0_reachability
,optional
)- VerifyReachability: Validate loopback reachability from the inband management VLAN interface.
- AvdTestLoopback0Reachability (Ansible tags:
loopback_reachability
,loopback0_reachability
)- VerifyReachability: Validate loopback reachability between devices.
- AvdTestLLDPTopology (Ansible tags:
lldp_topology
)- VerifyLLDPNeighbors: Validate LLDP topology.
- AvdTestMLAG (Ansible tags:
mlag
)- VerifyMlagStatus: Validate MLAG status.
- AvdTestRoutingTable (Ansible tags:
routing_table
)- VerifyRoutingTableEntry: Validate remote Loopback0 address and source interface for Vxlan1 interface are in the routing table.
- AvdTestBGP (Ansible tags:
bgp_check
)- VerifyBGPSpecificPeers: Validate IP BGP and BGP EVPN sessions state.
- VerifyRoutingProtocolModel: Validate ArBGP is configured and operating.
- AvdTestReloadCause (Ansible tags:
reload_cause
,optional
,never
)- VerifyReloadCause: Validate last reload cause. (Optional)
- (New) AvdTestAPIHttpsSSL (No Ansible tags, use the new
skipped_tests
variable instead)- VerifyAPIHttpsSSL: Validate eAPI HTTPS SSL profile status.
Input variables¶
# Root directory
root_dir: "{{ inventory_dir }}"
# AVD configurations output
# Main output directory
output_dir_name: "intended"
output_dir: "{{ root_dir }}/{{ output_dir_name }}"
# Output for test catalog YAML files if save_catalog is set to true
test_catalogs_dir_name: "test_catalogs"
test_catalogs_dir: "{{ output_dir }}/{{ test_catalogs_dir_name }}"
# Output directory for eos_validate_state reports
eos_validate_state_name: "reports"
eos_validate_state_dir: "{{ root_dir }}/{{ eos_validate_state_name }}"
# Output for test results JSON files if save_results is set to true
test_results_dir_name: "test_results"
test_results_dir: "{{ eos_validate_state_dir }}/{{ test_results_dir_name }}"
# Fabric name used in the reports name
fabric_name: "all"
# Reports name
eos_validate_state_md_report_path: "{{ eos_validate_state_dir }}/{{ fabric_name }}-state.md"
eos_validate_state_csv_report_path: "{{ eos_validate_state_dir }}/{{ fabric_name }}-state.csv"
# Input directory for custom ANTA catalogs:
custom_anta_catalogs_dir_name: "custom_anta_catalogs"
custom_anta_catalogs_dir: "{{ root_dir }}/{{ custom_anta_catalogs_dir_name }}"
# Allow different manufacturers
accepted_xcvr_manufacturers: "{{ validation_role.xcvr_own_manufacturers | arista.avd.default(['Arastra, Inc.', 'Arista Networks']) }}"
# Allow different states for power supplies
accepted_pwr_supply_states: "{{ validation_role.pwr_supply_states | arista.avd.default(['ok']) }}"
# Allow different states for fans
accepted_fan_states: "{{ validation_role.fan_states | arista.avd.default(['ok']) }}"
# Generate CSV results file
validation_report_csv: "{{ validation_role.validation_report_csv | arista.avd.default(true) }}"
# Generate MD results file
validation_report_md: "{{ validation_role.validation_report_md | arista.avd.default(true) }}"
# Print only FAILED tests
only_failed_tests: "{{ validation_role.only_failed_tests | arista.avd.default(false) }}"
# Variable to enable ANTA eos_validate_state
# Defaults to false as ANTA is currently in preview
use_anta: false
# Save each device test catalog to 'test_catalogs_dir'. Defaults to false.
save_catalog: false
# Logging level for the ANTA libraries. Defaults to "WARNING".
logging_level: "WARNING"
# The variable `skipped_tests` can be used for running/skipping test categories
skipped_tests:
- category: AvdTestHardware
# You can also decide to skip specific subtests (ANTA test name) for more granularity
skipped_tests:
- category: AvdTestBGP
tests:
- VerifyRoutingProtocolModel
- category: AvdTestHardware
tests:
- VerifyEnvironmentCooling
Custom ANTA catalog¶
You can now provide custom ANTA catalogs to the AVD eos_validate_state
role. By default, AVD will search for catalog YAML files in the custom_anta_catalogs
directory and incorporate these tests into the existing dynamically created catalog from AVD. The custom catalog files must be named as follows:
<hostname>.yml
or<hostname>.yaml
<group>.yml
or<group>.yaml
When specifying a group, it must be a group from the Ansible inventory. The custom tests will then be added to all devices that are part of this group. You can also use the all
group to target all the devices in your inventory. The directory where the custom catalogs are stored can be changed with the custom_anta_catalogs_dir
variable.
Warning
The skipped_tests
variable will ONLY skip the dynamically generated tests from the AVD validate state role. It will not skip tests added from custom catalogs.
Info
The final catalog will be validated by ANTA before running the tests on your network. Duplicate tests with the same inputs will be automatically removed. Therefore, dynamically generated tests by AVD will never be overwritten. To overwrite them, you should first skip them using the skipped_tests
variable and provide your own tests with inputs via a custom catalog.
Example Playbook¶
---
- name: validate states on EOS devices using ANTA
hosts: DC1
gather_facts: false
tasks:
- name: validate states on EOS devices
ansible.builtin.import_role:
name: arista.avd.eos_validate_state
vars:
# To enable ANTA
use_anta: true
# To save catalogs
save_catalog: true
Known issues¶
[__NSCFConstantString initialize] may have been in progress in another thread when fork() was called.
This issue affects OSX users only and is covered in Ansible documentation: https://docs.ansible.com/ansible/latest/reference_appendices/faq.html#running-on-macos-as-a-control-node.