The Pankian Metaphor 3126

operators were humans that did a certain set of operations with respect to application end .... but it was very structured subset ... since they weren't actually the humans directly responsible and-or necessarily really familiar with the application. with limited storage space ... the operators definitly were used to mount tapes and disks on the limited number of drives (i.e. there were significantly larger number of application tape reels and disk packs than there were drives).

i've often commented that one of the transitions for time-sharing services to 7x24 operations in the late 60s and early 70s was automating numerous operator functions ... and running the machine room "lights out" for 3rd and 4th shifts.

also, several years ago, the person running a major financial processing operation commented that the two major things responsible for 100percent availability over the past several years were

* ims hot-standby * automated operator

as hardware and software reliability has improved ... the major remaining source of faults were human (operator) mistakes and physical infrastructure failures and natural disasters. the ims hot-standby allowed for replicated operation at three geographically separated sites.

when my wife had been con'ed into doing a sting in POK in charge of loosely-coupled architecture ... she had developed "peer coupled shared data" architecture. there wasn't a lot of initial uptake until the ims hot-standby work and then later parallel sysplex stuff.

later when we were doing the high availability product (ha-cmp)

we had coined the terms disaster survivability and geographic survivability to differentiate from disaster-recovery.

--