If disconnect is called on a failover connection that is currently re-connecting, the call will hang until the connection is re-established.
This is just fine (and the expected behavior) when the connection is supposed to retry indefinitely (retry_attempts = 0). However, the same behavior is seen when retry_attempts is finite, and that's no good. In that scenario, once we have exceeded the retry attempt limit, we should bail on the disconnect process (unflushed buffers and all)
If
disconnect
is called on a failover connection that is currently re-connecting, the call will hang until the connection is re-established.This is just fine (and the expected behavior) when the connection is supposed to retry indefinitely (
retry_attempts = 0
). However, the same behavior is seen whenretry_attempts
is finite, and that's no good. In that scenario, once we have exceeded the retry attempt limit, we should bail on the disconnect process (unflushed buffers and all)