On Thu, 30 Jul 2015 05:43:43 -0500, Elardus Engelbrecht <[email protected]> wrote:
>Thomas Conley wrote: > >>I run into this all the time, especially when production, test, and sandbox >>LPARs are mixed into a sysplex. Typically the production systems share a >>RACF DB, the test systems share a different RACF DB, etc. Works fine. > >So, you have at least TWO shared RACF DBs in ONE SysPlex? At what z/OS or >OS/390 level? > >Beside that you mentioned in a later post that you set some bits to achieve >above, I'm not sure how you could do that, AFAIK, according to the RACF docs, >you can have only ONE shared RACF DB in ONE SysPlex. You are confusing several functions: (a) sharing the database, where you can have any number of different RACF databases shared among different combinations of systems in the sysplex, all controlled via their ICHRDSNT modules; (b) RACF Sysplex Communications, where all systems enabled for RACF sysplex communications (specified via ICHRDSNT) must use the same RACF database; (c) RACF Data Sharing (also specified via ICHRDST), where all the systems enabled for RACF sysplex communication may additionally share data and do some primitive communication via CF structures. Your statements about sharing (1 database) are true for systems in configurations (b) or (c), but not for systems in configuration (a). -- Walt ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
