Scope & goals:
After Software Deployment, DC/Cloud administrator would like to ensure the successful delivery of software & platform services as this gives them confidence that the infrastructure is ready to operate in production. The scope of this project was to develop an automated SDV(Software-Delivery-Validation) tool that validates the state of various software components across the cloud servers and checks health, configurations & settings of deployment. Further, generate a detailed report on the test results. Also, on the failure of any test-case collect extra information and logs so that end-user can look them later to debug further or pass this detailed report to Vendors. In summary, the initial goals of the SDV state validation tool were:
- Validating state, health-checks & configuration of the Delivered Software stack(Installer)
- Information gathering, collecting logs & report generation
Benefits: Operators & Vendors can debug faster for causes of bugs, deployment failures both in testing, production stage.
Tasks:
Week | Activity |
---|---|
Week 1 - Week 3 |
|
Week 4 - Week 6 |
|
Week 7 - Week 9 |
|
Week 10 - Week 12 |
|
Deliverables
- State Validation Tool Framework
- Airship deployment checks
- Tool-set documents
Evaluation Criteria
1st Evaluation (end of week 3): Pod Health Status Checker and Log collection
2nd Evaluation (end of week 6): Improve Airship deployment with custom scripts
3rd Evaluation (end of week 9): Software & Resource (Ceph, Monitoring/Logging agents health, Nova, CPU isolation, NTP, SR-IOV) checks
Final Evaluation (end of week 12): Automated Software State Validation tool and documentation
Methodology & Documentation:
Use OPNFV developer guideliness.
Outcomes:
Project is hosted inside CIRV-SDV repo in OPNFV Gerrit.
Demo: