The following message is a courtesy copy of an article that has been posted to bit.listserv.ibm-main,alt.folklore.computers as well.
[email protected] (Timothy Sipples) writes: > Linux on System z supports Discontinguous Saved Segments (DCSS), a z/VM > shared memory feature, and has for quite some time. There is a Linux block > device driver for DCSS (dcssblk.c) in the mainline Linux kernel > distribution. Features such as shared guest kernel(s), execute-in-place, > and certain monitoring functions take advantage of this capability, so it's > quite commonly used. Here's one source of information, for example: > > http://www.redbooks.ibm.com/abstracts/sg247285.html > > DCSS can be read/write (type SN) or read-only (type SR). re: http://www.garlic.com/~lynn/2009q.html#15 Mainframe running 1,500 Linux servers? I had done the original of DCSS as part of generalized paged mapped filesystem for CMS. Only a very small read-only subset of that was shipped in vm370 release 3. Old email discussing migrating the changes from cp67 to vm370 (early release 3). http://www.garlic.com/~lynn/2006v.html#email731212 http://www.garlic.com/~lynn/2006w.html#email750102 http://www.garlic.com/~lynn/2006w.html#email750430 none of the paged-mapped filesystem support was shipped (some A/B tests with modereate i/o thruput cms applications with optimized normal filesystem on 3380s ... ran avg. 3-times faster with the paged mapped filesystem ... some operations were significantly faster). misc. past posts http://www.garlic.com/~lynn/submain.html#mmap Part of demise of future system effort and rush to get stuff back into 370 product pipeline helped account for picking up bits & pieces of release. recent thread about other pieces going out as resource manager (another part of rushing things back into product pipeline, mvs/xa effort convinced corp. to kill vm370, shutdown the vm370 development effort, and move the people to POK to support mvs/xa ... in order to make the mvs/xa ship schedule; endicott managed to save the product mission ... but effectively had to reconstitute group from scratch). http://www.garlic.com/~lynn/2009q.html#14 Electric Light Orchestra IBM song, in 1981? a read-write subset (DWSS) was used later as part of original relational/sql implementation ... system/r ... misc. past posts mentioning system/r http://www.garlic.com/~lynn/submain.html#systemr and was supposed to have been part of tech. transfer of system/r to endicott for SQL/DS ... but DWSS changes were dropped before SQL/DS shipped and implementation had to be reworked to be done w/o it. for random other trivia ... one of the people mentioned in this jan92 meeting ... claimed to have handled the tech transfer back to STL for (mainframe) DB2 http://www.garlic.com/~lynn/95.html#13 some discussed in this recent post in comp.databases.theory http://www.garlic.com/~lynn/2009p.html#82 What would be a truly relational operating system ? http://www.garlic.com/~lynn/2009p.html#83 What would be a truly relational operating system ? and of course http://www.garlic.com/~lynn/2009p.html#43 From The Annals of Release No Sfotware Before Its Time http://www.garlic.com/~lynn/2009p.html#46 From The Annals of Release No Sfotware Before Its Time -- 40+yrs virtualization experience (since Jan68), online at home since Mar1970 ---------------------------------------------------------------------- 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
