An additional view of DBMS memory usage history. IMS OSAM data base buffer pools were a common pool shared by many data sets since its introduction. Originally only one block size was supported and most people chose a value like 6114 to obtain the best disk space utilization. The reason why IMS provided this service via its own access method (OSAM Overflow Sequencial Access Method originally meant to replace ISAM overflow) is an online system could support many data sets but only a few would be in active use assocated with currently scheduled programs at any moment in time.
I would like to interject (you will see why in a moment) that IMS and CICS were announced by IBM on the same day. Shortly after the introduction of VS operating systems that included VSAM something called the "VSAM DC OPTION" was introduced. The intended users were IMS and CICS. It had the same idea as the earlier common OSAM pool. The VSAM DC option first appeared in IMS/VS 1.0.1 along with a method to do block level sharing called Parallel DLI or PI (Program Isolation). Don't recall the CICS release number. The VSAM DC Option was renamed VSAM Shared Resources. There was a difference between the IMS and CICS deployment. IMS used "Global Shared Resources" meaning the VSAM pools were built in CSA. CICS used "Local Shared Resources" meaning the VSAM pools were built in private storage. IMS's pools were built in CSA because it always was multi address space and for speed cross address space SRBs were replaced with a lock / latch process. Now we get to the point of DBMS requirements driving changes in real and virtual memory archecture. In spite of several flavors of shared pools both IMS and CICS suffered from memory constraints. In the CICS world this was first addressed by trying to do what IMS did originally ... move out of the single address space model into Termanal owning, Application owning, and Data owning regions. In IMS (1.1.6 I think) LSO (Local Storage Option) was introduced where the VSAM buffers and the region connections were moved to there own address space called DLISAS. This movement was eventually enhanced by the development of cross memory services and there were 2 LSO options LSO=Y (Separate address space on machines that did not support cross memory) and LSO=S for machines that did support cross memory. With the introduction of IMS/VS 1.2 people stated to understand that even with IMS' natural multi region structure 1 IMS may not be enough. IMS were now being constrained by the private storage limitations of the LSO address space ... this was in 1982. The solution? IMS introduced a replacement for its old concept call Parallel DLI or PI (Program Isolation) with a new invention called Data Sharing and a new bit of software called the Inter-Region-Lock-Manager (some say IMS-Region-Lock-Manager) In classic fashion this new bit of software was refered to by its initials IRLM. Stepping back for a moment to IMS/VS 1.1.5 in 1979. IMS recovery was becoming very complex. New devices like the IBM MASS Storage device were introduced to support the very large amount of data that were being produced by DBMS's. IMS added a feature called DBRC or Data Base Recovery Control to both automate recovery and demonstrate the capabilities of Mass Storage. A few years latter IMS converted from Tape logging to DASD logging (IMS/VS 1.4 1984). >From a software architecture standpoint, Multi Regions, Independent locking (IRLM), Automated Recovery (DBRC), and DASD Logging became the foundations of IBMs second relational data base system and its first SQL based system called at its introduction DB2. IBMs first relational database system pre dated wide use of DASD and long historys could be written about it alone ... "Bill of Material Program called at various times BOMP T-BOMP for TAPE BOMP and D-BOMP for DISK BOMP. BOMP was probably used more for applications like payroll than manufacturing. The path of mapping changes in storage architecture to DBMS requirements is quite long and complicated. Hope this history from mostly the mid 70's to early 80's provided some in-site. Best Wishes Avram Friedman -----Original Message----- i've related before the discussion between IMS group and (original SQL/System/r group about pros & cons. http://www.garlic.com/~lynn/subtopic.html#systemr ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
