rel:: [[Profiling]] rel:: [[Performance|performance]] # Stochastic Profiling ## Causal Profiling From [Coz Paper](https://arxiv.org/pdf/1608.03676v1.pdf) > Unlike past profiling approaches, causal profiling indicates exactly where programmers should focus their optimization efforts, and quantifies their potential impact. Causal profiling works by running performance experiments during program execution. Each experiment calculates the impact of any potential optimization by virtually speeding up code: inserting pauses that slow down all other code running concurrently. The key insight is that this slowdown has the same relative effect as running that line faster, thus “virtually” speeding it up. Causal Profiling uses stochastic methods to determine bottlenecks in wall-clock response time throughput. ## Papers (from [Let The Computer Do The Work](x-devonthink-item://0B1E7C83-96D7-4881-AE92-0F1D0C2EBC8D) Article from Datadog Confluence wiki, Vector team, [original link](https://datadoghq.atlassian.net/wiki/spaces/VEC/pages/2126381424/Let+The+Computer+Do+the+Work)) - [Statistical Debugging for Real-World Performance Problems](https://dl.acm.org/doi/10.1145/2660193.2660234 "https://dl.acm.org/doi/10.1145/2660193.2660234") [PDF](http://petertsehsun.github.io/soen691/current/papers/oopsla161-song.pdf "http://petertsehsun.github.io/soen691/current/papers/oopsla161-song.pdf") - [SlowFuzz: Automated Domain-Independent Detection of Algorithmic Complexity Vulnerabilities](https://arxiv.org/abs/1708.08437 "https://arxiv.org/abs/1708.08437") [PDF](https://arxiv.org/pdf/1708.08437.pdf "https://arxiv.org/pdf/1708.08437.pdf") - [PerfFuzz: Automatically Generating Pathological Inputs](https://dl.acm.org/doi/10.1145/3213846.3213874 "https://dl.acm.org/doi/10.1145/3213846.3213874") [PDF](https://www.carolemieux.com/perffuzz-issta2018.pdf "https://www.carolemieux.com/perffuzz-issta2018.pdf") - [ConfProf: White-Box Performance Profiling of Configuration Options](https://dl.acm.org/doi/10.1145/3427921.3450255 "https://dl.acm.org/doi/10.1145/3427921.3450255") [PDF](https://software-lab.org/publications/icpe2021.pdf "https://software-lab.org/publications/icpe2021.pdf") - [Coz: finding code that counts with causal profiling](https://dl.acm.org/doi/10.1145/2815400.2815409 "https://dl.acm.org/doi/10.1145/2815400.2815409") [PDF](http://sigops.org/s/conferences/sosp/2015/current/2015-Monterey/printable/090-curtsinger.pdf "http://sigops.org/s/conferences/sosp/2015/current/2015-Monterey/printable/090-curtsinger.pdf") - [GAPP: A Fast Profiler for Detecting Serialization Bottlenecks in Parallel Linux Applications](https://arxiv.org/abs/2004.05628 "https://arxiv.org/abs/2004.05628") [PDF](https://arxiv.org/pdf/2004.05628.pdf "https://arxiv.org/pdf/2004.05628.pdf") - [[CRISP Critical Path Analysis of Large-Scale Microservice Architectures-Zhizhou Zhang]] ## Articles - [Let The Computer Do The Work](x-devonthink-item://0B1E7C83-96D7-4881-AE92-0F1D0C2EBC8D) - [Causal Profiling for Go](https://morsmachine.dk/causalprof)