This section describes various client-to-server and server-to-server connections.
Open Client applications establish a socket connection to Adaptive Server similarly to the way that existing client connections are established. Before any user data is transmitted, an SSL handshake occurs on the socket when the network transport-level connect call completes on the client side and the accept call completes on the server side.
Adaptive Server establishes a socket connection to another server for RPCs in the same way that existing RPC connections are established. Before any user data is transmitted, an SSL handshake occurs on the socket when the network transport-level connect call completes. If the server-to-server socket connection has already been established, then the existing socket connection and security context is reused.
When functioning as a client during RPCs, Adaptive Server requests the remote server’s certificate during connection. Adaptive Server then verifies that the CA that signed the remote server’s certificate is trusted; that is to say, on its own list of trusted CAs in the trusted roots file. It also verifies that the common name in the server certificate matches the common name used when establishing the connection.
You can use a companion server to configure Adaptive Server for failover. You must configure both the primary and secondary servers with the same SSL and RPC configuration. When connections fail over or fail back, security sessions are reestablished with the connections.
Component Integration Services, RepAgent, Distributed Transaction Management, and other modules in Adaptive Server use Client-Library to establish connections to servers other than Adaptive Server. The remote server is authenticated by its certificate. The remote server authenticates the Adaptive Server client connection for RPCs with user name and password.
