Client/server
Consul agents take part in failure detection and consensus to agree that a failure has occurred. Failure detection is done through periodic random probing. If a Consul agent in a failed node does not provide an acknowledgment, the other Consul agents are asked to probe the failed node. If no acknowledgment is received, the node is marked as suspicious,
but still remains a member of the cluster. The node is assumed dead if the suspicious node does not dispute its status within a configurable time period. Once a node is assumed dead, its status is gossiped to the entire cluster. HashiCorp has implemented modifications to the failure detection protocol and has achieved 20% faster failure detection with a 20 times reduction in false ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access