[Info-vax] VMS and the Internet of Things (IoT)

[Kerry Main]

> -----Original Message-----
> From: Info-vax [mailto:info-vax-bounces at rbnsn.com] On
> Behalf Of Jan-Erik Soderholm via Info-vax
> Sent: 11-Sep-16 7:03 AM
> To: info-vax at rbnsn.com
> Cc: Jan-Erik Soderholm <jan-erik.soderholm at telia.com>
> Subject: Re: [Info-vax] VMS and the Internet of Things
> (IoT)
> 
> Den 2016-09-11 kl. 12:16, skrev Simon Clubley:
> > I've seen a number of posts recently about VMS and
it's
> possible role
> > in an Internet of Things (IoT) environment.
> >
> > However, it appears to me that people are talking at
> cross-purposes
> > when talking about the IoT because the term itself is
so
> vaguely
> > defined and as such people appear to have their own
> differing and
> > unstated assumptions about what the IoT actually is.
> >
> > As such, it might be a good idea if people state what
> they mean by
> > the IoT and how they see the role that VMS has in it,
> because right
> > now I am not seeing a major role for VMS.
> >
> > Here's my initial pass at defining what I think the
IoT is all
> about
> > when I hear that phrase and why I think the above
> about VMS:
> >
> > I think of the IoT as being a three level architecture
with
> sensors
> > on devices at the lowest level, some kind of
coordinator
> or
> > controller within the facility at the medium level,
and
> remote
> > servers (if needed) at the highest level.
> >
> > At the lowest level, the sensors on devices level,
there's
> absolutely
> > no role for VMS at all. Most of these are going to be
> Cortex-M0/M4
> > level CPUs if even that as in some cases these might
just
> be small
> > 8-bit devices or even dumb sensors wired directly into
> the facility
> > controller.
> >
> > The medium level facility controller is where things
may
> start to get
> > more interesting but I don't see a role for VMS here
> either even if
> > you ignore that fact that VMS will not currently run
on
> the
> > architectures typically in use here.
> >

My references to OpenVMS and IoT were in respect to
OpenVMS and future architectures - coming (X86-64) and
potential (ARM). This would be post OpenVMS V9+ (ARM -
V10?) - after the new file system and new TCPIP stack.
X86-64 in small boxes/appliances are valid thin "smart"
clients just as much as lower power ARM devices in even
smaller thin client "smart" devices are likely. 

While there are some niche players in this thin smart
client market, the big OS to beat for OpenVMS on these
smaller devices is obviously Linux. Microsoft is on a
self-destruct model and is following in the DEC footsteps
of upsetting Customers and raising prices through the roof
at a time when Customers are under extreme pressures to
reduce IT costs. I see Microsoft share of the server
market dropping significantly in the future. Also,
OpenVMS/Linux can boot and run their latest versions in
1GB of memory (or less my Alpha servers boot latest
supported versions in 256MB) - try that with any version
of Windows Servers (or desktop).

110% pure speculation on my part, but I also suspect there
will be new licensing model options in OpenVMS V9+.

I stated "secure" thin client, because at some point, this
will need to happen. You cannot have driver-less cars or
smart traffic light systems being hacked. Unfortunately,
like what all too often happens, in the rush to get
products out the door and be first to market, the last
thing on developers/companies minds is security. As we all
know here, real security needs to be architected into the
entire solution - not added on later.

The more tiers, the higher risk of the secure chain being
breached. When "smart" thin clients are combined with
other technologies like IPV6 (low level thin client
devices can be identified by MAC addresses), it gets a lot
more interesting in terms of consolidating multiple
network tiers and returning more to a traditional secure
thin "smart" client and secure back ends. 

Yes, IPV6 has been slow to be adopted, and there will be
security issues with IPV6 to be addressed, but it is
coming. There are no other feasible options to replace
IPV4.

Btw -  as I stated in my earlier thread, this is a lot of
hype bundled in with this term IoT - you can literally
define it anyway you like. That's what everyone including
media analysts and vendors do anyway.

This is not unlike Public Clouds, OpenStack, Software
Defined Networks (SDN), and early industry hype of SOA
(granted, SOA does have technical validity, but not in the
way most of the industry hyped it for years i.e. SOA
everything).

[snip..]


Regards,

Kerry Main
Kerry dot main at starkgaming dot com