When you start questioning Linux terminology, the whole story starts to fall
apart. Hyperscale has meaning beyond marketing and programmers use it as to
boast about their skill. It happened to be the one that popped into my head. I
could have said bigdata, cloud, clusters or boat load of other terms that are
equally different to different people. The sad fact is that Linux has a lot of
terminology that average people are exposed to because there are so many
solutions to the same problem.
Maybe you should talk to a Google programmer about chunk locks for the Google
Filesystem that has been replaced by Google Collosus. Cobol programmers don't
need to know about chunk locks because it's hidden by update for a record,
Hyperscale is a configuration that meets high demands of shifting workload and
as you say is used a lot in reference to cloud computing. IBM can do this with
hardware and z/OS software whereas Linux is a software solution. In Linux,
maybe you do it by virtualization or by repurposing one or more computers. z
hardware and z/OS gives you a lot more options that are simpler (e.g. WLM,
LPAR/CPU defs, Hardware disk replication).
On Tuesday, July 18, 2023 at 07:20:47 PM PDT, Steve Thompson
<ste...@wkyr.net> wrote:
Linux in VM is a non-starter for hyperscale Linux computing. You might as well
by PC servers.
What is hyperscale Linux? I've been involved in LPARs of z/VM running multiple
Linux servers in SSI pairs for hot fail over. So I am curious what you mean by
this.
Looking this up, I read a lot of word salad, and salespeople/marketeers that
don't understand (or flatly ignore that Cloud is someone else's data center),
so their definitions are, really, word salad.
Steve Thompson
