Hazelcast CP Subsystem and Hazelcast’s failure detectors

Question

I am trying to configure the CP Subsystem in Hazelcast. Based on the Hazelcast documentation, it claims the following: "CP Subsystem relies on Hazelcast’s failure detectors to test the reachability of CP members. Before removing a CP member from CP Subsystem, please make sure that it is declared as unreachable by Hazelcast’s failure detector and removed from Hazelcast cluster’s member list." (Here is the source: https://docs.hazelcast.com/imdg/4.0/cp-subsystem/management)

Could someone provide useful information regarding how one could inject the Failure Detector in a Spring Boot application in order to ensure this statement? Or what is the way to ensure that the CP member is unreachable by Hazelcast's failure detector?

Thanks in advance!

Answer 1

Hazelcast comes with configured deadline failure detector out of the box. See documentation. The maximum period without heartbeat is rather conservative 60 s, used to be 120 s. The documentation points out that you should not remove the CP member during this time period before cluster correctly detects the member dead.

Other failure detector options are

• Phi Accrual Failure Detector
• Ping Failure Detector