If tcp_retries2 is exhausted the transport's socket is transitioned to TCP_CLOSED without sending FIN,RST. That's perfectly fine behavior until you have a middle-box that silently drops the re-connection's SYNs because it keeps an established state around until torn down by a FIN or RST. NFSv4 won't usually make it to tcp_retries2 since it sets RPC_CLNT_CREATE_DISCRTRY to tear the socket down after a minor timeout, thus avoiding this particular problem with this particularly stupid middle-box. Can anyone tell me if there is a reason not to set RPC_CLNT_CREATE_DISCRTRY for NFSv3, or make it configurable? Ben -- To unsubscribe from this list: send the line "unsubscribe linux-nfs" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
