Continuous Profiling, the missing piece in your observability puzzle!
summary
Continuous profiling claims to run in production with low overhead so teams catch regressions and save cloud cost, using eBPF-based collectors or language agents like Java Flight Recorder and storing continuous samples for comparison. OpenTelemetry adds profiling hooks; open source options include Periscope Pixie Parka. Google’s 2010 design underpins modern systems; practitioners report operational wins (production visibility, proactive fixes) and typical overheads around <2%. Storage and symbolization remain the main pain points.
AI-written summary · about
chapters (18)
- 0:00Introduction
- 1:48Monoliths to microservices
- 6:05Distributed systems challenges
- 7:27Three pillars of observability
- 10:01Traditional profiling limits
- 13:14Profiling as observability layer
- 16:49Profiling in production limits
- 18:12What is continuous profiling
- 20:20Modern profiling architecture
- 23:53Key profiling challenges
- 25:17Cost savings and performance
- 27:25Proactive incident detection
- 28:11Profile acquisition and bias
- 31:04eBPF low-overhead profiling
- 33:52Open source profiling tools
- 36:48eBPF unified collection
- 38:13OpenTelemetry profiling future
- 39:16Q&A
related talks
01h 13mContinuous Profiling: The Missing Piece in Your Observability Puzzle
52mThe missing piece in your observability puzzle!
56mContinuous Profiling the missing piece in your observability puzzle
46mContinuous Profiling - the 4th Pillar of Observability
47mContinuous Profiling — what, why and how
24mWhy Does Continuous Profiling Matter to Developers?
24mCommunity Driven Breakout Session: Continuous Profiling
36mProfiling in the cloud-native era - Matthias Loibl