>From a dispatching perspective, there is 0 cost. The idle TCBs are not on any work-eligible queues.
>From a system perspective, there is a non-0 cost for things such as -- Setting all the tasks non-dispatchable (as happens if taking an SVC Dump of your space). That doesn't take much per idle TCB, but it's more than 0. -- Repairing the system (what we call "global recovery", for what that's worth) - each work unit must be evaluated Neither of these should contribute to any decision about the number of TCBs you choose. There are obvious storage limits related to the limitations of (for example) below-16M storage for TCBs and RBs in the address space. Peter Relson z/OS Core Technology Design ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
