In a cluster mode setup of EventStoreDB, when the leader node of the cluster drops and we attempt to use ReadStream method, it raises an error and doesn’t rebalance with the updated cluster members.
If we call ReadStream again, it fails repeatedly because client.grpcClient.handleError is not invoked in case of an error, which prevents the reconnection process from being triggered.
In a cluster mode setup of EventStoreDB, when the leader node of the cluster drops and we attempt to use
ReadStream
method, it raises an error and doesn’t rebalance with the updated cluster members.This issue occurs here: https://github.com/EventStore/EventStore-Client-Go/blob/8206ac84067e5a6d591f71fafb82758015fb9f3c/esdb/client.go#L291
If we call ReadStream again, it fails repeatedly because client.grpcClient.handleError is not invoked in case of an error, which prevents the reconnection process from being triggered.
Here’s a suggested solution https://github.com/EventStore/EventStore-Client-Go/pull/188