Keeping data safe in the presence of crashes is a fundamental problem storage systems. Although high-level ideas for crash consistency are relatively well understood, realizing them practice surprisingly complex and full challenges. The systems research community actively working on solving this challenge, papers examined here offer three solutions.