Fuzzing Nomad with Nomad
summary
Shows a reproducible stack that runs go-fuzz harnesses on HashiCorp Nomad using Terraform/Packer on Google Cloud, with Prometheus+Grafana for metrics and a small custom go-fuzz exporter that parses go-fuzz stderr. Jobs register as services so Prometheus discovers metrics through Nomad job with Connect sidecar. Example details: docker image built, tmpfs mount with 4GB, go-fuzz defaults to all CPU cores; experiments show multi-core raises executions/sec but coverage converges with a single-core cap. Also demonstrates a distributed coordinator+workers setup and that the pipeline successfully finds real crashes.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 0:32Fuzzing basics and Golang
- 12:05Nomad infrastructure stack
- 21:02Observability for fuzzing
- 23:15Fuzzing Nomad with Nomad
- 31:15Q&A and conclusion
related talks
40mNomad: Scheduling Applications at Scale
39mFuzzing Session: Finding Bugs and Vulnerabilities Automatically
44mFuzzing in Go: Fear, Despair, Acceptance and the Long-Awaited Success
24mGetting More Confident with Your Security Helper Libraries Thanks to Go Fuzzing
53mFuzz Testing for Fun and Profit
01h 00mClusterFuzz: Fuzzing at Google Scale
37mDistributed Flexibility: Nomad and Vault in a Post-Kubernetes World
44mFuzzing in Go: Fear, Despair, Acceptance and Long-Awaited Success ☠️