Summary: from ACME to ROC
21st Century Research challenge is Availability, Changability, Maintainability, Evolutionary Growth
Peres’s Law vs. Moore’s Law: Cope with fact that people, SW, HW fail
Industry: may soon compete on recovery v. SPEC
Need theory for design of Internet services
Recovery Oriented Computing is one path
- Failure data collection + Benchmarks to evaluate ACME
- Partitioning, Redundancy, Input Insertion, Diagnosis
- Undo (3 R’s), Fast Restart lowers MTTR
- Margin of Safety via More 9s in target?
Significantly reducing MTTR (people/SW/HW) => better Availability & Cost of Ownership