Grokking the System Design Interview, Volume II
Vote

0% completed

​

11. Phi Accrual Failure Detection

Basic heartbeating only tells you a node is up or down, with no signal in between. The Phi Accrual Failure Detector replaces that boolean with a suspicion level that rises as evidence of failure builds.

Background

In distributed systems, detecting failure is genuinely hard. A system that responds slowly under heavy load or network congestion can look the same as one that is down. There is no certain way to tell the two apart.

Heartbeating is the usual answer, and it only reports a boolean: alive or not alive. It works off a fixed timeout

.....

.....

.....

Like the course? Get enrolled and start learning!

Reading Progress

0%

Vote for new content