当前位置:首页 > 报告详情

用于 OpenBMC 的 AI 辅助单元测试和集成测试框架.pdf

上传人: 明**** 编号:1011920 2025-12-21 17页 892.09KB

1、Manish Baing,LenovoArun Mahendran,LenovoAI assist Unit and Integration test framework for OpenBMCAI assist Unit and Integration test framework for OpenBMCManish Baing,LenovoArun Mahendran,LenovoTest and ValidationBackgroundProblem statement Test framework Architecture Modular Unit Testing Fault inje

2、ctions Integration testing AI assist test cases generation to improve Code Coverage AI assist Cyclomatic Complexity computationCall to Action OutlineBackgroundTesting requirements continue to change at each stage from POC to MP.The variety of server CPU architectures and server platform components w

3、ill result in multiple use cases with unique requirements,requiring specific hardware configurations for each testing phase.Baseboard Management controller(BMC)is present in most of servers motherboard.It enables remote monitoring of health and availability of server infrastructure.The OpenBMC is a

4、Linux Foundation open-source project that produces an open-source implementation of the BMC firmware stack.BMC/OpenBMCThe complexity involved in simulating and injecting hardware failure scenarios,such as Power Supply Unit(PSU)faults,Add-in card(Storage,Graphics,Network)failures,CPU and Memory error

5、s,Fan failures etc.poses significant challenges for OpenBMC validation.Limited availability of faulty hardware&the high cost of physical test setups,including specific hardware,worsen these issues.Module-based code development increases the risk of missing existing or domain-specific functional scen

6、arios during new code development.Problem StatementModule level Unit testingStub CreationSimplified Wrapper and Interface Generator(SWIG)Python UnittestFault injection Integration testingReports&Alerts Deployment Overview of Test frameworkA stub is interface which simulates a real object.Here Real o

word格式文档无特别注明外均可编辑修改,预览文件经过压缩,下载原文更清晰!
三个皮匠报告文库所有资源均是客户上传分享,仅供网友学习交流,未经上传用户书面授权,请勿作商用。
主要内容概括: 1. **背景**:服务器测试需求不断变化,硬件故障模拟复杂,物理测试成本高,模块化代码开发存在风险。 2. **问题陈述**:需要模块级单元测试、故障注入、集成测试等。 3. **测试框架架构**:使用SWIG生成接口,进行模块化单元测试,通过Python unittest进行测试。 4. **AI辅助**:利用AI生成测试用例,提高代码覆盖率,计算Cyclomatic Complexity。 5. **挑战与目标**:优化框架,使其与OpenBMC架构兼容,提高性能和可扩展性。
"AI助力OpenBMC测试,效率翻倍?" "如何用AI提升OpenBMC代码覆盖率?" "OpenBMC测试框架,AI能做什么?"
客服
商务合作
小程序
服务号
折叠