Introduction

Introduction
Prev	Chapter 13. GT.CM Client/Server	Next

GT.CM is the network client/server database software for GT.M. GT.CM on UNIX allows access to GT.M databases residing on a server, by client processes running on multiple nodes on a network.

GT.CM consists of a Server and a Client. The Server is a network object that performs database operations on behalf of GT.M Client processes, running on other nodes of the network. GT.CM uses TCP task-to-task communication facilities for the link between a client GT.M process and a server. GT.CM on UNIX operates properly between supported platforms independent of the native byte ordering The GT.CM client is packaged within the GT.M run-time system.

When a GT.M process requires access to a database on another node, it sends a request across the network to a GT.CM Server process running on that node, and the Server accesses the database. The process requesting access to a database through GT.CM is referred to as a Client process. The node from which the data is requested is referred to as the Server node.

The use of GT.CM is largely transparent to the GT.M application code. The only visible change in the application.s environment is the addition of error messages delivered in case of problems in network operations.

GT.CM can communicate between systems having different endian architectures.

	Notes
GT.M transaction processing (TP) is not supported via GT.CM, and accessing GT.CM served databases within an M TRANSACTION may cause application level inconsistency in the data. GT.CM servers do not invoke triggers. This means that the client processes must restrict themselves to updates which don't require triggers, or explicitly call for the actions that triggers would otherwise perform. Because GT.CM bypasses triggers, it may provide a mechanism to bypass triggers for debugging or complex corrections to repair data placed in an inconsistent state by a bug in trigger logic.

Notes

GT.M transaction processing (TP) is not supported via GT.CM, and accessing GT.CM served databases within an M TRANSACTION may cause application level inconsistency in the data.
GT.CM servers do not invoke triggers. This means that the client processes must restrict themselves to updates which don't require triggers, or explicitly call for the actions that triggers would otherwise perform. Because GT.CM bypasses triggers, it may provide a mechanism to bypass triggers for debugging or complex corrections to repair data placed in an inconsistent state by a bug in trigger logic.

In the event of recovery from system crashes, application level database consistency cannot be guaranteed for data residing in databases (M global variable namespaces) accessed via different GT.CM servers or distributed between GT.CM and local access.