Observability and Control Theory
Submitted by Piyush Verma (@meson10) on Sunday, 18 August 2019
Section: Full talk (40 mins) Category: SRE
Software, by default, is opaque. To see what it’s doing you have to write observation capability into it. This goes beyond logs and stepping through in a debugger - because you have to observe the live system, not your sandbox. Is observability a new concept and how is it different from monitoring?
How does Control theory use observability to build systems that thrive on feedback and improve? How TCP has been using Control Theory and observability forever to build reliable packet delivery, that allows you to read this text.
- Sample problem that go unnoticed with traditional Blackbox monitoring
- Sample problem that goes unnoticed with advanced whitebox monitoring.
- Timeline of Monitoring tools.
- Evolution of Monitoring needs.
- What is monitoring vs Observability?
- Servers vs.Services.
- Real questions that Distributed Systems have to answer
- Domains of Observability
- Need of Observability
- Debuggability in advanced systems.
- Maths of Stability
- What is Control theory?
- Practicale example of Control Theory
- Types of Control-Systems
- How Control theory uses Observability
- How do we use Control theory
- How TCP Implements all of this together
- FLow Control vs Congestion Control
- Sample implementation in a working application
- Thank you?
Piyush Verma heads Site Reliabilty Engineering at TrustingSocial.com. At present, their systems span across 6 datcenters and 3 major clouds across 3 countries.
He has been working with Infrastrcuture Problems for over a decade and in past life he has built datscale.io, oogway.in, and siminars.com.