OpenTP1 Version 7 TP1/Client User's Guide TP1/Client/W, TP1/Client/P
This section provides notes on communication errors in TP1/Client and how they are handled.
- Output of an error message when a communication error occurs
TP1/Client uses TPC, a connection-type communication protocol, and implements it with a socket interface. TP1/Client therefore detects communication errors as system call errors at the sockets interface. If a system call error occurs, TP1/Client outputs an error message to identify the error cause.
Error messages are output to error log files. For details on error log files, see 2.11.1 Error logging.
- Communication error related to ports
TP1/Client establishes and releases a connection each time communication occurs. If multiple CUPs repeatedly issue the dc_clt_cltin_s function and dc_clt_cltout_s function or issue processing in succession that repeatedly performs the dc_rpc_call_s function, an error may occur because the ports of the operating system are all temporarily being used. If such an error occurs, increase the number of ports to tune the setting to a value acceptable for operation. Alternatively, wait until ports are released and ready for operation, and then retry the communication.
- Communication error that occurs when the dc_rpc_call_s function is used
If the dc_rpc_call_s function is issued while the communication destination schedule service is starting or terminating, the system switches to another host to perform the processing. If one of the hosts specified in the target_host argument of the dc_clt_cltin_s function or DCHOST in the client environment definition has not started, the dc_rpc_call_s function may require some time until it returns.
If the dc_rpc_call_s function is issued during establishment of a permanent connection or within a transaction, the host is not switched when an error response is received.
All Rights Reserved. Copyright (C) 2006, 2009, Hitachi, Ltd.