Currently a node is unhealthy for a fixed 30 minutes. Nodes in Hedera can be very fragile. Treating every issue as a node being down for 30 minutes can cause the whole network to come down quickly in a client. Other SDKs start at 250ms and backoff based on successive ticks of "being unhealthy" up to an hour.
Problem
Currently a node is unhealthy for a fixed 30 minutes. Nodes in Hedera can be very fragile. Treating every issue as a node being down for 30 minutes can cause the whole network to come down quickly in a client. Other SDKs start at 250ms and backoff based on successive ticks of "being unhealthy" up to an hour.
Solution
Use a backoff instead of a fixed timeout.
Alternatives
No response