Labels

Continuous Diagnostics and Mitigation (US Civ gov Cybersecurity program) 100 OEMs, $1 Billion/year

IPS/IDS program for Internet access run by Verizon, ATT, CenturyLink, 2nd largest Cyber spend by Civ gov

Civilian, State, and Local gov Highly Adaptive Cybersecurity Services program for Threat Hunting (Pen testing, Risk assessments, high value assessments) services

Civilian gov mandated high value asset governance tied to CDM with Quarterly reports to Congress and Comptroller General

centralized IPS/IDS program for DoD/Allies coordinated by DISA, 37+ OEMs in the Joint Regional Security Stack. So big and awkwardly implemented it's drawn formal Pentagon criticism in 2017 and 2018. Multi-billion dollar IPS/IDS.

Unhealthy test issues with 399 or more flakiness reports.

Unhealthy test issues with between 37 and 398 flakiness reports.

Unhealthy test issues with between 13 and 36 flakiness reports.

Unhealthy test issues with less than 12 flakiness reports.

When a flaky test issue was reviewed (mainly use by Developer Experience for triaging purposes)

The test assumes the dataset is in a particular (usually limited) state, which might not be true depending on when the test run during the test suite.

The test is assuming a specific date or time.

A flaky test issue arising from timing-related factors, such as delays, eventual consistency, asynchronous operations, or race conditions. These issues may stem from shortcomings in the test logic, the system under test, or their interaction. While tests can sometimes address these issues through improved synchronization, they may also reveal underlying system bugs that require resolution.

The test use random values, that sometimes match the expectations, and sometimes not.

Test data state leaked through test cases. See https://docs.gitlab.com/ee/development/testing_guide/flaky_tests.html#state-leak

Master broken incidents caused by flaky tests which are failing because sql queries limit has exceeded

Test is failing intermittently due to a transient bug in our software

The DOM selector used in the test is unreliable.

The test fails from time to time due to infrastructure issues.

Follow up issues that are created from merge requests