The database and a few other variables maintained by each stack entity are defined as global definitions. This has no implication in OS LIKE Unix, which do not share memory ACROSS tasks. However, in case of RTOS (like Vxworks, PSOS) which have a flat memory space, there is only one copy of this DATA irrespective of NUMBER of tasks or processes started. e.g. if two instances of ISUP are running on the same board they will be referring to the same copy of data.

This will be fixed in future releases of stack – the solution will be as decided in the Protocol Development framework, which would be one of the following :

• Keep global database indexed by the INSTANCE number of the module.
• Define the database as local to the stack entity and pass the reference to it with each call of stack and down to all levels. All the database references will be done with respect to this reference.

The database and a few other variables maintained by each stack entity are defined as global definitions. This has no implication in OS like Unix, which do not share memory across tasks. However, in case of RTOS (like Vxworks, PSOS) which have a flat memory space, there is only one copy of this data irrespective of number of tasks or processes started. e.g. if two instances of ISUP are running on the same board they will be referring to the same copy of data.

This will be fixed in future releases of stack – the solution will be as decided in the Protocol Development framework, which would be one of the following :