Heartbeat detection between clusters

"Royal" Hangzhou-sunface () 2015/11/23 Monday a.m. 10:24:04erl-kernel Net_ticktime or NET_KERNEL:SET_NET_TICKTIME/1 can set the live heartbeat interval between clusters "Royal" Hangzhou-sunface () 2015/11/23 Monday 10:24:31 If the node is less, you can consider setting it short, strengthening the real-time of server down monitoring "local tyrants" Lotus-Four Seasons () 2015/11/23 Monday morning 10:27:07 What if it  's shortened, timed out, and didn't return to the heartbeat? "Royal" Hangzhou-sunface () 2015/11/23 Monday 10:28:01 with a ticktime/4 interval to monitor, so if 4 times are not successful, it is considered that the cluster's peer node is down, then the node will be removed from the cluster "local tyrants" Lotus Next-Four seasons (2015/11/23) Monday 10:29:01 Oh, my. "Royal" Hangzhou-sunface () 2015/11/23 Monday 10:30:05 Of course not to say remove, is marked down, if good can continue to join the cluster of "local tyrants" lotus-Four Seasons meditation () 2015/11/23 Monday 10:30:56 en, The supervisor will then pull the off-group node back. "Royal" Hangzhou-sunface () 2015/11/23 Monday 10:31:32 It's supposed to be that node. If the heartbeat packet is issued, the cluster is flagged as live after it is received

Heartbeat detection between clusters