On Mon, 31 May 2010 17:13:27 -0400, Shmuel Metz (Seymour J.) wrote: >In <listserv%201005240741293308.0...@bama.ua.edu>, on 05/24/2010 > at 07:41 AM, Paul Gilmartin aid: > >>ISPF also does ENQ SYSDSN SHR, and for load modules, SYSIEWLP, so it >>provides protection from programs not doing the same thing. > >No, only from programs allocating exclusive. It doesn't protect from >programs using DISP=SHR without the ISPF coordination. > Programs updating PDS members using DISP=SHR provide an integrity exposure if other programs do likewise unless some additional serialization mechanism is used. What's necessary is that all programs updating PDS members use identical (or at least coordinated) serialization mechanisms. In that sense, relying on ISPF coordination introduces no new exposure in environments not already using another convention.
(I'm not saying this well.) If a site uses only DISP=SHR, but relies on job name serialization (all jobs updating a given PDS have identical jobnames), then, yes, introducing reliance on ISPF serialization with a different job name (or TSO session) introduces a hazard. ISPF serialization solved an outstanding and pervasive availability problem. It should be adopted more widely. In the case of the C RTL, the programmer should be given the choice of ENQ SYSDSN EXC or ISPF serialization which has an excellent track record when used properly. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@bama.ua.edu with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
