This document provides the release notes for Hunter of Doctor.

Important notes

In Hunter release, Doctor has been working with fault management alarming enhancement and maintenance use case to upstream.

OpenStack Fenix (unofficial project) has been futher worked to implement the infrastructure rolling maintenance in interaction with VNFM. Doctor CI testing also supports maintenance testing with Fenix latest master version.

MCP and Apex installers are currently supported. In MCP we test fault management without the enhanced alarm. In Apex we test fault management with alarm enhancement with sample implementation. Apex is also used for maintenance use case testing. Congress inspector support is not working.

In this release, Doctor has not yet been working to look outside OpenStack. We might need to get back to this later to better meet the tough industry requirements as well as requirements from edge, containers and 5G.

Summary

Hunter Doctor framework uses OpenStack Rocky integrated into its test cases. For testing, we use Doctor CI on OPNFV installers.

Release Data

Doctor changes

commit-ID Subject
b19b69d731cfb5a87f7c928cc898ea04ec85cec6 Fix tox to clean python cache
364d2c2344be5775a0eef6fe19fda125d2c8853d Handle the exception for running the profiler
e6c857ba931a03fb7bfb49746cc4a7eb5b6ce6e1 Hunter release documentation
73605c5c34b97ab56306bfa9af0f5888f3c7e46d Support Fenix as admin tool
33293e9c23a21ad3228f46d2063f18c915eb2b79 Wrong yamllint disable command
d82ab34f15a9b67185c85c6afc5562bc8b72cb8b Add local.conf.sample for devstack deployment of Doctor
7ecc40b445b2aa42f0680c96dc672accf4e40ba0 Hi, the automation job failed.
2cd1ca4463121e2354fd920af2b26c65848fb9e2 Remove Ryota from committers
f31ab961c594595772b0c3d4bd40a0d9491fc6cb Removing committers
c653d95c67436698296e238396bf5d8370e3169a Update to INFO file
e6708c869855ab69f9b53959befd82bb2f32f9ad Bug - Testing in Apex with OpenStack master fails
e1c5dd0158d5168738fcc9918d24c04ca724b056 remove to set ceilometer config in MCP
d673e9218a53e047edc5ff6cd047ac6db5112651 Support Fenix and sample implementation accordingly
916e4931a56c1a5d41d46148609bf348d4326d37 fix the configparser for Python 2 and 3 Compatibility
4075b417e973adb257ae39ff5c25aa182a2af3ea index.rst was blank
dafdfcfad6866d7c413d4b8d5a9d25f3ab1f76dc Minor docs updates

Releng changes

commit-ID Subject
cc290b2f937a2edbd60a5d2d1e20f333dfc7eb88 Doctor to run Fenix as admin tool
488c558492201aacd359305a7afa3d5640a90b0e Add parameter of DEPLOY_SCENARIO for doctor
30478e1e193485cce93164e9877002b811acf950 remove SSH_KEY parameter from doctor-slave-parameter
513b05275cbac2ff98950bb0a384a275dd8884f5 Parpare ssh_key for MCP in doctor

Version change

Module version changes

  • OpenStack has changed from Queens-1 to Rocky-1 since previous Gambia release.

Document version changes

These documents have been updated in Hunter release

  • Testing document docs/development/overview/testing.rst
  • Doctor scenario in functest: docs/development/overview/functest_scenario/doctor-scenario-in-functest.rst

Reason for version

Documentation is updated due to maintenance use case testing using Fenix.

Feature additions

JIRA REFERENCE SLOGAN
DOCTOR-129 Maintenance use case implementation in OpenStack Fenix
DOCTOR-131 Support Fenix and sample implementation accordingly
DOCTOR-132 Integrate with MCP
DOCTOR-133 Doctor fault management with notification from Inspector

Deliverables

Software deliverables

None

Documentation deliverables

https://git.opnfv.org/doctor/tree/docs

Known Limitations, Issues and Workarounds

System Limitations

Maintenance test case requirements:

  • Minimum number of nodes: 1 Controller, 3 Computes
  • Min number of VCPUs: 2 VCPUs for each compute

Known issues

None

Workarounds

None

Test Result

Doctor CI results with TEST_CASE=’fault_management’ and INSPECTOR_TYPE=sample

TEST-SUITE Results:
INSTALLER_TYPE=’Apex’ 1) SUCCESS
INSTALLER_TYPE=’Fuel’ 2) SUCCESS
INSTALLER_TYPE=’Local’ N/A
  1. Uses enhanced alarming worked in DOCTOR-133
  2. Uses alarm from Nova reset server state API generated notification. API call can take a lot of time and alarm might take over a second. This also happens with different installer and for more Telco grade performance we need the enhanced alarming.

Doctor CI results with TEST_CASE=’fault_management’ and INSPECTOR_TYPE=congress

TEST-SUITE Results:
INSTALLER_TYPE=’Apex’ 1) FAILED
INSTALLER_TYPE=’Fuel’ N/A
INSTALLER_TYPE=’Local’ N/A
  1. Takes over one second because Nova reset server state error API is too slow

Doctor Functest results with TEST_CASE=’fault_management’

TEST-SUITE Results:
INSTALLER_TYPE=’Apex’ skipped
INSTALLER_TYPE=’Fuel’ skipped
INSTALLER_TYPE=’Local’ N/A

Note: Installer Functest does not currently test features or skips running the project test cases

Doctor CI results with TEST_CASE=’maintenance’

TEST-SUITE Results:
INSTALLER_TYPE=’Apex’ ADMIN_TOOL_TYPE=’sample’ SUCCESS
INSTALLER_TYPE=’Apex’ ADMIN_TOOL_TYPE=’fenix’ SUCCESS

Doctor Functest results with TEST_CASE=’maintenance’

N/A - Needs special target environment with at least 3 compute nodes

References

For more information about the OPNFV Doctor latest work, please see:

https://wiki.opnfv.org/display/doctor/Doctor+Home