What? -------------- Original message from Anne & Lynn Wheeler <[EMAIL PROTECTED]>: --------------
> 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] (Ted MacNEIL) writes: > > Mainframers brag about how long a system stays up. PFCSK's brag about > > how fast they can re-boot. > > a lot of work was done on cp67 and subsequent vm370 to stay up, reboot > quickly, and require minimum hands-on for 7x24 operation for > around-the-clock, online access. > > reboot quickly included backup and operation in lights-out, unattended > operation. > > this helped allow cp67 to migrate into commercial time-sharing service. > first shift was obviously, fairly heavily used ... but offshift and > weekends use might be problematical ... making it difficult to justify > offshift availability unless costs were reduced to a minimum (i.e. no > operator). however, there were other characteristics that were needed. > back then, there was high percentage of leased systems ... and the > processor "meter" was used to establish monthly charges. in addition to > work on unattended/automated operation for cp67 ... there was also work > on how to get the meter to "stop" when the system was otherwise idle. > the system meter would run when the processor was executing and/or when > there was active i/o operations (and would coast for 400mills after > things quiesced). it took a little slight of hand to leave active i/o on > the (terminal) interface (able to responds to spontaneous terminal i/o) > and not have the meter run. > > for other topic drift ... posts related to getting to play disk engineer > in bldgs. 14 (disk engineering) and 15 (disk product test). at the time > they had tried operating test machines (processors used for testing > disks & controllers in development) under MVS ... but found it had a > MTBF of 15 minutes (hang and/or crash ... both requiring reboot). > http://www.garlic.com/~lynn/subtopic.html#disk > > sort of as a hobby, putting together bullet proof operations where they > could not only operate a single "test cell" ... but operate several > concurrently (eliminated scheduling bottleneck for dedicated stand-alone > single test at a time). > > ---------------------------------------------------------------------- > 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 > ---------------------------------------------------------------------- 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