code analysis - Howto overcome Unit Test Regression Problems...? -
i looking kind of solution software development teams spend time handling unit test regression problems (about 30% of time in case!!!), i.e., dealing unit tests fails on day day basis.
following 1 solution i'm familiar with, analyzes of latest code changes caused unit test fail:
unit test regression analysis tool
i wanted know if knows similar tools can benchmark them. well, if can recommand approach handle annoying problem.
thanks @ advanced
you have our sympathy. sounds have brittle test syndrome. ideally, single change unit test should break single test-- , should real problem. said, "ideally". type of behavior common , treatable.
i recommend spending time team doing root cause analysis of why these tests breaking. yep, there fancy tools keep track of tests fail often, , ones fail together. continuous integration servers have built in. that's great. suspect if ask each other, you'll know. i've been though , team knows experience.
anywho, few other things i've seen cause this:
- unit tests shouldn't depend on more class , method testing. dependencies have crept in. make sure you're using dependency injection make testing easier.
- are these unique tests? or testing same thing on , over? if going fail together, why not remove one?
- many people favor integration on unit tests, since more coverage buck. these, single change can break lots of tests. maybe you're writing integration tests?
- perhaps running through common set-up code lots of tests, causing them break in unison. maybe can mocked out isolate behaviors.
Comments
Post a Comment