Dear Boss
In a three-node environment,all three stolon-proxy nodes cannot connect to stolon-keeper,and the error message is
not proxy to master address since aren't in the enabled proxies list {"address":"177,0,0,56:5432"}
master address {"address":"177,0,0,56:5432"}
there is continuously printed.
is is possible to design a strategy that reports an error after failing to eonnect etcd a certain number of times, instead of
diaplaying it as an info level log? in this way,the pod will be restarted and the service can continue to be used.
Compared to data consistency, availability may be more importtant. Alternatively, can we cache the write logs of a certain
time period after reporting an error, so that developers can recover them later?
Dear Boss In a three-node environment,all three stolon-proxy nodes cannot connect to stolon-keeper,and the error message is not proxy to master address since aren't in the enabled proxies list {"address":"177,0,0,56:5432"} master address {"address":"177,0,0,56:5432"} there is continuously printed.
is is possible to design a strategy that reports an error after failing to eonnect etcd a certain number of times, instead of diaplaying it as an info level log? in this way,the pod will be restarted and the service can continue to be used. Compared to data consistency, availability may be more importtant. Alternatively, can we cache the write logs of a certain time period after reporting an error, so that developers can recover them later?
thanks !
k8s+three node+3 keeper +3 proxy + 3 sentinel +etcd as backend