Discussion:
What is the future role for VMS ?
(too old to reply)
Simon Clubley
2017-06-06 18:25:17 UTC
Permalink
Raw Message
Since it's come up once again, it's time to ask people outright
what market segments you see as being a possible future for VMS.

Some questions to get you started:

Is x86-64 VMS a server OS for both new and existing users ?

Is x86-64 VMS a server OS for existing users only ?

Is x86-64 VMS a desktop OS (at least for some people) ?

Is x86-64 VMS some kind of embedded/process control OS ?

My answers:

I think the only possible future for VMS for the next 5+ years is
going to be for existing users looking to keep VMS applications
running within their organisation.

Past that, you are going to need some unique application or feature
VMS can offer which would be unique enough that it would persuade
new people to try it and VMS.

However, I do not currently see what that can be.

Do you see such a possible unique application or feature ?

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world
Arne Vajhøj
2017-06-06 18:38:40 UTC
Permalink
Raw Message
Post by Simon Clubley
Since it's come up once again, it's time to ask people outright
what market segments you see as being a possible future for VMS.
Is x86-64 VMS a server OS for both new and existing users ?
This is best case.

Getting new users will ensure that VMS has a viable future.
Post by Simon Clubley
Is x86-64 VMS a server OS for existing users only ?
This is worst case.

Easy to achieve but the slow natural attrition will kill VMS in the long
run.
Post by Simon Clubley
Is x86-64 VMS a desktop OS (at least for some people) ?
No chance assuming a mainstream definition of what a desktop OS
provide.

A few VMS lovers may run VMS on a desktop system, but it is not a
market.
Post by Simon Clubley
Is x86-64 VMS some kind of embedded/process control OS ?
I don't think so.

This is a very crowded market.
Post by Simon Clubley
I think the only possible future for VMS for the next 5+ years is
going to be for existing users looking to keep VMS applications
running within their organisation.
Past that, you are going to need some unique application or feature
VMS can offer which would be unique enough that it would persuade
new people to try it and VMS.
However, I do not currently see what that can be.
Do you see such a possible unique application or feature ?
Difficult to see. OS'es are becoming commodities.

They could consider selling VMS as risk reduction through
diversity.

Arne
Simon Clubley
2017-06-07 18:25:11 UTC
Permalink
Raw Message
Post by Arne Vajhøj
Post by Simon Clubley
Since it's come up once again, it's time to ask people outright
what market segments you see as being a possible future for VMS.
Is x86-64 VMS a server OS for both new and existing users ?
This is best case.
Yes, it is.
Post by Arne Vajhøj
Getting new users will ensure that VMS has a viable future.
This can also be sub-divided into new users who were VMS customers
in the past and new users who have never used VMS.
Post by Arne Vajhøj
Post by Simon Clubley
Is x86-64 VMS a server OS for existing users only ?
This is worst case.
Easy to achieve but the slow natural attrition will kill VMS in the long
run.
Also agreed. This is the "managed decline" scenario.

VMS is certainly not going anywhere near the desktop market and any
possible options in the embedded/process control world (such as they
are, given all the existing options out there) are limited to the
higher levels of such an infrastructure.
Post by Arne Vajhøj
They could consider selling VMS as risk reduction through
diversity.
Now _that's_ an interesting possibility - to try selling VMS by using
the "monocultures are bad for security" argument.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world
j***@yahoo.co.uk
2017-06-07 23:03:54 UTC
Permalink
Raw Message
Post by Simon Clubley
Post by Arne Vajhøj
Post by Simon Clubley
Since it's come up once again, it's time to ask people outright
what market segments you see as being a possible future for VMS.
Is x86-64 VMS a server OS for both new and existing users ?
This is best case.
Yes, it is.
Post by Arne Vajhøj
Getting new users will ensure that VMS has a viable future.
This can also be sub-divided into new users who were VMS customers
in the past and new users who have never used VMS.
Post by Arne Vajhøj
Post by Simon Clubley
Is x86-64 VMS a server OS for existing users only ?
This is worst case.
Easy to achieve but the slow natural attrition will kill VMS in the long
run.
Also agreed. This is the "managed decline" scenario.
VMS is certainly not going anywhere near the desktop market and any
possible options in the embedded/process control world (such as they
are, given all the existing options out there) are limited to the
higher levels of such an infrastructure.
Post by Arne Vajhøj
They could consider selling VMS as risk reduction through
diversity.
Now _that's_ an interesting possibility - to try selling VMS by using
the "monocultures are bad for security" argument.
Simon.
--
Microsoft: Bringing you 1980s technology to a 21st century world
Wrt monocultures: there's a reason "dissimilar redundancy"
came into existence. It was originally 'robustness' rather
than what's now called 'security', but the two are somewhat
related.

There's quite a lot written about dissimilar redundancy e.g.
in life-critical systems on board the International Space
Station, but the core concept goes back at least as far as
the last century, in things like fly by wire architectures.

Last time I looked, it was hard to find any real volume
examples of where it's still in common use. That could
change, if people were motivated to care, and the return
on investment was good enough.
j***@gmail.com
2017-06-07 02:18:37 UTC
Permalink
Raw Message
On occasion, I have Googled VMS, and most of the hits are related to virtual machines (VMs). So, as long as VSI has the operating system source code, how about this: While supporting the existing customer base who would like to execute their VMS applications on commodity hardware, turn VMS into the ultimate VM hosting system. It actually has an advantage in that it does not resemble the systems that are most often hosted as VMs -- Windows and Linux. It could be configured as a dedicated VM hosting platform.

The above is somewhat tongue-in-cheek, and I have no idea if it is feasible or how you could possibly improve on the features of VMware, the convenience of VirtualBox, etc., etc., but if you're trying to invent a niche . . .

j
Arne Vajhøj
2017-06-07 04:03:14 UTC
Permalink
Raw Message
Post by j***@gmail.com
On occasion, I have Googled VMS, and most of the hits are related to
virtual machines (VMs). So, as long as VSI has the operating system
source code, how about this: While supporting the existing customer
base who would like to execute their VMS applications on commodity
hardware, turn VMS into the ultimate VM hosting system. It actually
has an advantage in that it does not resemble the systems that are
most often hosted as VMs -- Windows and Linux. It could be
configured as a dedicated VM hosting platform.
I think most production use bare metal hypervisors.

Arne
Kerry Main
2017-06-10 15:05:41 UTC
Permalink
Raw Message
-----Original Message-----
Vajhøj via Info-vax
Sent: June 7, 2017 12:03 AM
Subject: Re: [Info-vax] What is the future role for VMS ?
Post by j***@gmail.com
On occasion, I have Googled VMS, and most of the hits are related to
virtual machines (VMs). So, as long as VSI has the operating system
source code, how about this: While supporting the existing customer
base who would like to execute their VMS applications on commodity
hardware, turn VMS into the ultimate VM hosting system. It actually
has an advantage in that it does not resemble the systems that are
most often hosted as VMs -- Windows and Linux. It could be
configured as a dedicated VM hosting platform.
I think most production use bare metal hypervisors.
Arne
A bare metal hypervisor is nothing more than another marketing name for
a native clustered OS (albeit customized for VM hosting) which, in
VMware's case, is called ESXi.

You could call a clustered OpenVMS environment a bare metal hypervisor
as well.

In fact, VMware is an active-passive cluster where clients can only
connect to one VM on that hypervisor, so a bare metal hypervisor called
OpenVMS which presented an active-active load balanced hosting
capability across multiples sites might be an attractive alternative at
some point in the future.

Now, VMware does have some impressive capabilities, like dynamic
migration of VM's between servers, but in VMware's case, the entire VM
exists in a single file which makes it easier to migrate. Imho, that
would be something to look at for future OpenVMS versions.


Regards,

Kerry Main
Kerry dot main at starkgaming dot com
Arne Vajhøj
2017-06-12 17:00:41 UTC
Permalink
Raw Message
Post by Kerry Main
Vajhøj via Info-vax
Post by j***@gmail.com
On occasion, I have Googled VMS, and most of the hits are related to
virtual machines (VMs). So, as long as VSI has the operating system
source code, how about this: While supporting the existing customer
base who would like to execute their VMS applications on commodity
hardware, turn VMS into the ultimate VM hosting system. It actually
has an advantage in that it does not resemble the systems that are
most often hosted as VMs -- Windows and Linux. It could be
configured as a dedicated VM hosting platform.
I think most production use bare metal hypervisors.
A bare metal hypervisor is nothing more than another marketing name for
a native clustered OS (albeit customized for VM hosting)
A bare metal hypervisor has nothing to do with a cluster.

A bare metal hypervisor manage VM's that are running utilizing the
hardwares virtualization support.

A cluster makes a number of nodes collaborate on doing some tasks.

The two concepts are completely orthogonal.
Post by Kerry Main
You could call a clustered OpenVMS environment a bare metal hypervisor
as well.
No.

VMS could provide a host hypervisor. But given that VMS is an OS and
not a CPU then it can not provide a bare metal hypervisor. VMS can run
on a bare metal hypervsior.
Post by Kerry Main
In fact, VMware is an active-passive cluster where clients can only
connect to one VM on that hypervisor,
????

VMWare can host both independent VM's and VM's in cluster
both active-active and active-passive. VMWare does not care
about that - that is all up to the OS and the applications.

Arne
Kerry Main
2017-06-17 14:41:48 UTC
Permalink
Raw Message
-----Original Message-----
Vajhøj via Info-vax
Sent: June 12, 2017 1:01 PM
Subject: Re: [Info-vax] What is the future role for VMS ?
Arne
Post by Kerry Main
Vajhøj via Info-vax
Post by j***@gmail.com
On occasion, I have Googled VMS, and most of the hits are related
to
Post by Kerry Main
Post by j***@gmail.com
virtual machines (VMs). So, as long as VSI has the operating system
source code, how about this: While supporting the existing
customer
Post by Kerry Main
Post by j***@gmail.com
base who would like to execute their VMS applications on
commodity
Post by Kerry Main
Post by j***@gmail.com
hardware, turn VMS into the ultimate VM hosting system. It actually
has an advantage in that it does not resemble the systems that are
most often hosted as VMs -- Windows and Linux. It could be
configured as a dedicated VM hosting platform.
I think most production use bare metal hypervisors.
A bare metal hypervisor is nothing more than another marketing name
for
Post by Kerry Main
a native clustered OS (albeit customized for VM hosting)
A bare metal hypervisor has nothing to do with a cluster.
A bare metal hypervisor manage VM's that are running utilizing the
hardwares virtualization support.
A VMware hypervisor is a collection of servers that each require a dedicated NIC on each server for cluster communications (VMotion traffic). The cluster "applications" are simply multiple containers that have embedded OS instances.
A cluster makes a number of nodes collaborate on doing some tasks.
The two concepts are completely orthogonal.
Again, a VMware farm requires a dedicated cluster communication NIC to handle its inter-node communications.
Post by Kerry Main
You could call a clustered OpenVMS environment a bare metal
hypervisor
Post by Kerry Main
as well.
No.
VMS could provide a host hypervisor. But given that VMS is an OS and
not a CPU then it can not provide a bare metal hypervisor. VMS can run
on a bare metal hypervsior.
Its all in the marketing.

ESXi is a customized clustered Linux OS that focusses on running multiple "containers" called VM's across multiple servers. These containers have the ability to use virtual NIC's to provide the ability to be able to act as independent OS's.

Similar to what Galaxy OS's can do as well.

To its credit, the VMware cluster can dynamically relocate these containers to other nodes in the cluster in the event of a server failure. It does this because each container (VM) is just a single file on the SAN (shades of LD!).
Post by Kerry Main
In fact, VMware is an active-passive cluster where clients can only
connect to one VM on that hypervisor,
????
VMWare can host both independent VM's and VM's in cluster
both active-active and active-passive. VMWare does not care
about that - that is all up to the OS and the applications.
Arne
That part is correct. I should have stated that the vast majority of VMware clients (primarily Windows/Linux) only connect to one VM on the hypervisor.

If running Windows/Linux, they usually state that if a server fails in a VMware farm, that VM container will be restarted on a another server in the cluster / hypervisor and in their view, that is "good enough". Hence the reason why many Windows/Linux VM's are not clustered on top of a VMware "farm".


Regards,

Kerry Main
Kerry dot main at starkgaming dot com
Kerry Main
2017-06-10 15:22:55 UTC
Permalink
Raw Message
-----Original Message-----
Sent: June 10, 2017 11:06 AM
Subject: RE: [Info-vax] What is the future role for VMS ?
-----Original Message-----
Arne
Vajhøj via Info-vax
Sent: June 7, 2017 12:03 AM
Subject: Re: [Info-vax] What is the future role for VMS ?
Post by j***@gmail.com
On occasion, I have Googled VMS, and most of the hits are related to
virtual machines (VMs). So, as long as VSI has the operating system
source code, how about this: While supporting the existing customer
base who would like to execute their VMS applications on commodity
hardware, turn VMS into the ultimate VM hosting system. It actually
has an advantage in that it does not resemble the systems that are
most often hosted as VMs -- Windows and Linux. It could be
configured as a dedicated VM hosting platform.
I think most production use bare metal hypervisors.
Arne
A bare metal hypervisor is nothing more than another marketing name
for
a native clustered OS (albeit customized for VM hosting) which, in
VMware's case, is called ESXi.
You could call a clustered OpenVMS environment a bare metal
hypervisor
as well.
In fact, VMware is an active-passive cluster where clients can only
connect to one VM on that hypervisor, so a bare metal hypervisor
called
OpenVMS which presented an active-active load balanced hosting
capability across multiples sites might be an attractive alternative
at
some point in the future.
Now, VMware does have some impressive capabilities, like dynamic
migration of VM's between servers, but in VMware's case, the entire VM
exists in a single file which makes it easier to migrate. Imho, that
would be something to look at for future OpenVMS versions.
To clarify my previous reply, the differentiator for OpenVMS would be
dynamic process (not system/VM) migration between different systems,
with Galaxy type capacity planning business rule features that provide
dynamic add/delete of cpu's between different OS instances, high speed
cluster/network interconnects via shared memory or RoCEv2(?) on the same
big servers etc.


Regards,

Kerry Main
Kerry dot main at starkgaming dot com
IanD
2017-06-12 06:46:18 UTC
Permalink
Raw Message
Post by Kerry Main
-----Original Message-----
<snip>
Post by Kerry Main
To clarify my previous reply, the differentiator for OpenVMS would be
dynamic process (not system/VM) migration between different systems,
with Galaxy type capacity planning business rule features that provide
dynamic add/delete of cpu's between different OS instances, high speed
cluster/network interconnects via shared memory or RoCEv2(?) on the same
big servers etc.
Regards,
Kerry Main
Kerry dot main at starkgaming dot com
I see a real differentiator as being a system where a process can be kept infinitely alive and migrated anywhere

If you computing a huge workload that run for say 2 days and then the system dies, what good is a cluster?

I want to be able to shadow my process n x's for as much redundancy as I wish

Hadoop gets around this by farming out the workload into chunks with replication so that if a process dies others take over but they are not atomic in nature, if all the chunks dies, the processes have to start again from scratch

This is a poor mans cluster but when hardware is cheap and your spinning up 100's - 1000's of nodes, who cares

VMS clusters need to evolve and offer more than they do. Fully fault tolerant processes that can be migrated around a cluster and keep going would be an evolution that perhaps some companies may pay for?
David Froble
2017-06-12 16:34:34 UTC
Permalink
Raw Message
Post by IanD
Post by Kerry Main
-----Original Message-----
<snip>
Post by Kerry Main
To clarify my previous reply, the differentiator for OpenVMS would be
dynamic process (not system/VM) migration between different systems,
with Galaxy type capacity planning business rule features that provide
dynamic add/delete of cpu's between different OS instances, high speed
cluster/network interconnects via shared memory or RoCEv2(?) on the same
big servers etc.
Regards,
Kerry Main
Kerry dot main at starkgaming dot com
I see a real differentiator as being a system where a process can be kept
infinitely alive and migrated anywhere
Heck, even people need to sleep, sometime ..
Post by IanD
If you computing a huge workload that run for say 2 days and then the system
dies, what good is a cluster?
How many workloads would have such conditions?
Post by IanD
I want to be able to shadow my process n x's for as much redundancy as I wish
Hadoop gets around this by farming out the workload into chunks with
replication so that if a process dies others take over but they are not
atomic in nature, if all the chunks dies, the processes have to start again
from scratch
This is a poor mans cluster but when hardware is cheap and your spinning up
100's - 1000's of nodes, who cares
VMS clusters need to evolve and offer more than they do. Fully fault tolerant
processes that can be migrated around a cluster and keep going would be an
evolution that perhaps some companies may pay for?
The big question is, how many potential customers, and how much would they pay?

As an example, in the environment I'm currently working in, transactions are
done in less than a second, or at most a few seconds.

Now, I'm aware that there can be needs for disaster tolerant systems, and that
there are environments that can have workloads that can need much longer run
times for a task. A VMS cluster can fit the first requirement, not the second
so easily.

I've never use a VMS cluster. Never had the need. Maybe came close a few
times, but when costs were explored the customer would decide they didn't need
the extra capabilities. Usually the cost of hardware wasn't the issue, it was
the cost of high speed dedicated communications.
Arne Vajhøj
2017-06-12 16:46:50 UTC
Permalink
Raw Message
Post by David Froble
Post by IanD
Post by Kerry Main
To clarify my previous reply, the differentiator for OpenVMS would be
dynamic process (not system/VM) migration between different systems,
with Galaxy type capacity planning business rule features that provide
dynamic add/delete of cpu's between different OS instances, high speed
cluster/network interconnects via shared memory or RoCEv2(?) on the same
big servers etc.
I see a real differentiator as being a system where a process can be kept
infinitely alive and migrated anywhere
If you computing a huge workload that run for say 2 days and then the system
dies, what good is a cluster?
How many workloads would have such conditions?
Post by IanD
Hadoop gets around this by farming out the workload into chunks with
replication so that if a process dies others take over but they are not
atomic in nature, if all the chunks dies, the processes have to start again
from scratch
This is a poor mans cluster but when hardware is cheap and your spinning up
100's - 1000's of nodes, who cares
VMS clusters need to evolve and offer more than they do. Fully fault tolerant
processes that can be migrated around a cluster and keep going would be an
evolution that perhaps some companies may pay for?
The big question is, how many potential customers, and how much would they pay?
This may be a niche market. But it would likely be a market where
buyers would be willing to pay a premium.

I would say that it is worth considering.

I like the idea.

Biggest question is whether HPE will allow it. I think it will
somewhat be targeting the NSK market!
Post by David Froble
As an example, in the environment I'm currently working in, transactions
are done in less than a second, or at most a few seconds.
Sure. But I don't think VMS capabilities should be limited to what
everybody needs.

Arne
David Froble
2017-06-12 19:26:00 UTC
Permalink
Raw Message
Post by Arne Vajhøj
Post by David Froble
Post by IanD
Post by Kerry Main
To clarify my previous reply, the differentiator for OpenVMS would be
dynamic process (not system/VM) migration between different systems,
with Galaxy type capacity planning business rule features that provide
dynamic add/delete of cpu's between different OS instances, high speed
cluster/network interconnects via shared memory or RoCEv2(?) on the same
big servers etc.
I see a real differentiator as being a system where a process can be kept
infinitely alive and migrated anywhere
If you computing a huge workload that run for say 2 days and then the system
dies, what good is a cluster?
How many workloads would have such conditions?
Post by IanD
Hadoop gets around this by farming out the workload into chunks with
replication so that if a process dies others take over but they are not
atomic in nature, if all the chunks dies, the processes have to start again
from scratch
This is a poor mans cluster but when hardware is cheap and your spinning up
100's - 1000's of nodes, who cares
VMS clusters need to evolve and offer more than they do. Fully fault tolerant
processes that can be migrated around a cluster and keep going would be an
evolution that perhaps some companies may pay for?
The big question is, how many potential customers, and how much would they pay?
This may be a niche market. But it would likely be a market where
buyers would be willing to pay a premium.
I would say that it is worth considering.
I like the idea.
Indeed! I like it. My only concern is it's commercial viability, and what else
might suffer from the resources devoted to such.
Post by Arne Vajhøj
Biggest question is whether HPE will allow it. I think it will
somewhat be targeting the NSK market!
Don't know that HPE could have much to say about it. Anything VSI developes
belongs to VSI, as far as I know, and I doubt VSI would have agreed to HPE
taking back anything granted to VSI.
Post by Arne Vajhøj
Post by David Froble
As an example, in the environment I'm currently working in,
transactions are done in less than a second, or at most a few seconds.
Sure. But I don't think VMS capabilities should be limited to what
everybody needs.
True. One of the strengths of VMS is how it can be used in so many environments.
Arne Vajhøj
2017-06-12 20:10:34 UTC
Permalink
Raw Message
Post by David Froble
Post by Arne Vajhøj
Post by David Froble
The big question is, how many potential customers, and how much would they pay?
This may be a niche market. But it would likely be a market where
buyers would be willing to pay a premium.
I would say that it is worth considering.
I like the idea.
Indeed! I like it. My only concern is it's commercial viability, and
what else might suffer from the resources devoted to such.
Post by Arne Vajhøj
Biggest question is whether HPE will allow it. I think it will
somewhat be targeting the NSK market!
Don't know that HPE could have much to say about it. Anything VSI
developes belongs to VSI, as far as I know, and I doubt VSI would have
agreed to HPE taking back anything granted to VSI.
There may be some competition clauses in the HPE-VSI agreement or
VSI may have a business need to stay on friendly terms with HPE.

Arne
Kerry Main
2017-06-17 14:13:35 UTC
Permalink
Raw Message
-----Original Message-----
Vajhøj via Info-vax
Sent: June 12, 2017 12:47 PM
Subject: Re: [Info-vax] What is the future role for VMS ?
Post by David Froble
Post by IanD
Post by Kerry Main
To clarify my previous reply, the differentiator for OpenVMS would
be
Post by David Froble
Post by IanD
Post by Kerry Main
dynamic process (not system/VM) migration between different
systems,
Post by David Froble
Post by IanD
Post by Kerry Main
with Galaxy type capacity planning business rule features that
provide
Post by David Froble
Post by IanD
Post by Kerry Main
dynamic add/delete of cpu's between different OS instances, high
speed
Post by David Froble
Post by IanD
Post by Kerry Main
cluster/network interconnects via shared memory or RoCEv2(?) on
the same
Post by David Froble
Post by IanD
Post by Kerry Main
big servers etc.
I see a real differentiator as being a system where a process can
be
kept
Post by David Froble
Post by IanD
infinitely alive and migrated anywhere
If you computing a huge workload that run for say 2 days and then
the
Post by David Froble
Post by IanD
system
dies, what good is a cluster?
How many workloads would have such conditions?
Post by IanD
Hadoop gets around this by farming out the workload into chunks
with
Post by David Froble
Post by IanD
replication so that if a process dies others take over but they are not
atomic in nature, if all the chunks dies, the processes have to
start
Post by David Froble
Post by IanD
again
from scratch
This is a poor mans cluster but when hardware is cheap and your spinning up
100's - 1000's of nodes, who cares
VMS clusters need to evolve and offer more than they do. Fully
fault
Post by David Froble
Post by IanD
tolerant
processes that can be migrated around a cluster and keep going
would
Post by David Froble
Post by IanD
be an
evolution that perhaps some companies may pay for?
The big question is, how many potential customers, and how much
would
Post by David Froble
they pay?
This may be a niche market. But it would likely be a market where
buyers would be willing to pay a premium.
I would say that it is worth considering.
I like the idea.
Biggest question is whether HPE will allow it. I think it will
somewhat be targeting the NSK market!
NonStop achieves its fault tolerance at the SW level - not the HW level.
This has been true since NonStop moved to Itanium and Intel removed the
functions from Itanium that NonStop required to do this at the HW level.

Fwiw, I do not think HPE is interested in OS platforms anymore. It has
given up, sold off its SW business and is returning to its core focus as
an infrastructure/server/storage provider with an enterprise consulting
capability. This also makes HPE ripe for a take-over, but that is
another discussion. I expect HPE will spin-off NonStop in a manner
similar to the OpenVMS/VSI model.
Post by David Froble
As an example, in the environment I'm currently working in,
transactions
Post by David Froble
are done in less than a second, or at most a few seconds.
Sure. But I don't think VMS capabilities should be limited to what
everybody needs.
Arne
Agree 100%.

Like the vehicle market, there are lots of different requirements for
compute platforms. OpenVMS is being positioned as a solid server
platform, so it needs to be able to address a multitude of back end
server requirements.

Having stated this, I do not agree that fault tolerance is one of those
areas that would provide the payback or added value differentiator to
the majority of the server Cust's today.

There is a trade-off between very expensive fault tolerance and very
high HA. It’s a bit like the difference between a Tier 3 DC and one that
is Tier4. Almost no one builds Tier 4 DC's anymore (except Govt types
maybe), because it is orders of magnitudes cheaper and with more
flexibility to build a couple of Tier 3's in pairs that are less than
100km apart configured as a VDC (virtual DC).

Stock exchanges, likely candidates for fault tolerance are evolving.
They now need very high HA - not fault tolerance. Their need for speed
has in some cases over ridden the need for fault tolerance. Hence, it is
not uncommon to have a stock exchange running in active-passive mode
between sites which means that in the event of a 9/11 type incident,
they WILL lose data. The amount of data lost would depend on the
replication timeframe between site updates and when the incident
occurred.

In some cases, you might see a stock exchange say something like "we
gained a lot of speed with our new system!". What they will not say is
that they also gave up their RPO=0 requirement.

Hence, this is why I see the NonStop market continuing to decline.


Regards,

Kerry Main
Kerry dot main at starkgaming dot com
Simon Clubley
2017-06-18 09:16:41 UTC
Permalink
Raw Message
Post by Kerry Main
Stock exchanges, likely candidates for fault tolerance are evolving.
They now need very high HA - not fault tolerance. Their need for speed
has in some cases over ridden the need for fault tolerance. Hence, it is
not uncommon to have a stock exchange running in active-passive mode
between sites which means that in the event of a 9/11 type incident,
they WILL lose data. The amount of data lost would depend on the
replication timeframe between site updates and when the incident
occurred.
Of course, that worked really well for BA... :-)
Post by Kerry Main
In some cases, you might see a stock exchange say something like "we
gained a lot of speed with our new system!". What they will not say is
that they also gave up their RPO=0 requirement.
Is giving up RPO=0 a false economy ?

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world
Kerry Main
2017-06-18 12:32:14 UTC
Permalink
Raw Message
-----Original Message-----
Simon Clubley via Info-vax
Sent: June 18, 2017 5:17 AM
Subject: Re: [Info-vax] What is the future role for VMS ?
Post by Kerry Main
Stock exchanges, likely candidates for fault tolerance are evolving.
They now need very high HA - not fault tolerance. Their need for speed
has in some cases over ridden the need for fault tolerance. Hence, it is
not uncommon to have a stock exchange running in active-passive
mode
Post by Kerry Main
between sites which means that in the event of a 9/11 type incident,
they WILL lose data. The amount of data lost would depend on the
replication timeframe between site updates and when the incident
occurred.
Of course, that worked really well for BA... :-)
Post by Kerry Main
In some cases, you might see a stock exchange say something like "we
gained a lot of speed with our new system!". What they will not say is
that they also gave up their RPO=0 requirement.
Is giving up RPO=0 a false economy ?
Simon.
Given the nature of the business, losing data may be an acceptable risk
given the alternative costs to prevent such an occurrence.

If a single transaction might contain millions of $'s, or loss of life,
then it may not be an acceptable risk.

Each business needs to determine how much DR/DT "insurance" they are
willing to pay for to address the basic risks, while accepting that if a
significant event does occur, they may be in deep trouble.

Unfortunately, most businesses today are increasingly depending on IT.

BA found this out the hard way.

Another example - company went bankrupt when they did not follow DR/DT
and security best practices with offsite backups separate from the
cloud.

<http://www.computerworld.com/article/2491008/cloud-security/hacker-puts
--full-redundancy--code-hosting-firm-out-of-business.html>
"CodeSpaces.com shut down after a hacker gained access to its Amazon EC2
account and deleted most data, including backups"
"A code-hosting and project management services provider was forced to
shut down operations indefinitely after a hacker broke into its cloud
infrastructure and deleted customer data, including most of the
company's backups."

With off-site backups, they would have still been impacted, but they
would at least been able to recover.

Regards,

Kerry Main
Kerry dot main at starkgaming dot com
Jan-Erik Soderholm
2017-06-18 14:00:33 UTC
Permalink
Raw Message
Post by Kerry Main
-----Original Message----- From: Info-vax
[Info-vax] What is the future role for VMS ?
Post by Kerry Main
Stock exchanges, likely candidates for fault tolerance are
evolving. They now need very high HA - not fault tolerance. Their
need for
speed
Post by Kerry Main
has in some cases over ridden the need for fault tolerance. Hence,
it is
Post by Kerry Main
not uncommon to have a stock exchange running in active-passive
mode
Post by Kerry Main
between sites which means that in the event of a 9/11 type
incident, they WILL lose data. The amount of data lost would depend
on the replication timeframe between site updates and when the
incident occurred.
Of course, that worked really well for BA... :-)
Post by Kerry Main
In some cases, you might see a stock exchange say something like
"we gained a lot of speed with our new system!". What they will not
say
is
Post by Kerry Main
that they also gave up their RPO=0 requirement.
Is giving up RPO=0 a false economy ?
Simon.
Given the nature of the business, losing data may be an acceptable risk
given the alternative costs to prevent such an occurrence.
If a single transaction might contain millions of $'s, or loss of life,
then it may not be an acceptable risk.
Each business needs to determine how much DR/DT "insurance" they are
willing to pay for to address the basic risks, while accepting that if
a significant event does occur, they may be in deep trouble.
Unfortunately, most businesses today are increasingly depending on IT.
I think I spotted a slight mistyping there. One "Un-" too much... :-)

I guess that most of us here on this forum, are just happy that
their custumers are using IT more and more.
Simon Clubley
2017-06-18 15:20:34 UTC
Permalink
Raw Message
Post by Jan-Erik Soderholm
Post by Kerry Main
Each business needs to determine how much DR/DT "insurance" they are
willing to pay for to address the basic risks, while accepting that if
a significant event does occur, they may be in deep trouble.
Unfortunately, most businesses today are increasingly depending on IT.
I think I spotted a slight mistyping there. One "Un-" too much... :-)
I guess that most of us here on this forum, are just happy that
their custumers are using IT more and more.
I wonder how many people reading comp.os.vms were impacted by the
BA failure, the NHS shutdown or the various banking systems failures ?

IOW, it might be nice as a source of employment but what happens
when you or others are affected personally because those IT systems
were not sufficiently robust (BA/banks) or had undisclosed
vulnerabilities (NHS) ?

Kerry's right to say "unfortunately", provided that word is used in the
context of not having sufficient robustness/redundancy built into those
same systems.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world
Simon Clubley
2017-06-18 15:13:27 UTC
Permalink
Raw Message
Post by Kerry Main
Given the nature of the business, losing data may be an acceptable risk
given the alternative costs to prevent such an occurrence.
The problem with that is that the person who takes that decision
may not be the person who has to deal with the consequences of
that decision.

As such it's all too easy to wrongly convince yourself that a certain
level of redundancy is not required, especially if you have a nicely
formatted consultant's report you can quote from to "justify" your
decision.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world
David Froble
2017-06-18 19:56:13 UTC
Permalink
Raw Message
Post by Simon Clubley
Post by Kerry Main
Given the nature of the business, losing data may be an acceptable risk
given the alternative costs to prevent such an occurrence.
The problem with that is that the person who takes that decision
may not be the person who has to deal with the consequences of
that decision.
So, you're talking with the guy that gets a bonus depending on the next
quarter's profit, and has a guaranteed golden parachute?
Post by Simon Clubley
As such it's all too easy to wrongly convince yourself that a certain
level of redundancy is not required, especially if you have a nicely
formatted consultant's report you can quote from to "justify" your
decision.
In my opinion, there is never enough redundancy.

Don't trust tapes ...
Don't trust disks ...
Haven't decided about paper ...
Paul Sture
2017-06-18 15:21:15 UTC
Permalink
Raw Message
Post by Kerry Main
-----Original Message-----
Simon Clubley via Info-vax
Is giving up RPO=0 a false economy ?
Simon.
Given the nature of the business, losing data may be an acceptable risk
given the alternative costs to prevent such an occurrence.
If a single transaction might contain millions of $'s, or loss of life,
then it may not be an acceptable risk.
No publicly quoted company can afford to lose financial transactions, no
matter how small; if the auditors refuse to sign off the accounts, the
whole company is toast.
Post by Kerry Main
Each business needs to determine how much DR/DT "insurance" they are
willing to pay for to address the basic risks, while accepting that if
a significant event does occur, they may be in deep trouble.
Unfortunately, most businesses today are increasingly depending on IT.
BA found this out the hard way.
One niggling question about BA's disaster - various commentators were
questioning that if the cause was an upgrade gone wrong (there was a
Tweet which suggested that), why did they do it on a UK holiday weekend
when so many passengers might be affected?

One possible answer to that is that management decided that avoiding a
disruption to Business Class passengers during the week was of a higher
priority. There are different ways to alleviate risks - at one place we
decided that any upgrade taking less than an hour or two happened over a
lunch time, so that in the event of something going wrong, all the
application teams were on hand to diagnose and fix the problem(s). :-)
--
Everybody has a testing environment. Some people are lucky enough to
have a totally separate environment to run production in.
Kerry Main
2017-06-19 00:24:17 UTC
Permalink
Raw Message
-----Original Message-----
Sture via Info-vax
Sent: June 18, 2017 11:21 AM
Subject: Re: [Info-vax] What is the future role for VMS ?
Post by Kerry Main
-----Original Message-----
Simon Clubley via Info-vax
Is giving up RPO=0 a false economy ?
Simon.
Given the nature of the business, losing data may be an acceptable risk
given the alternative costs to prevent such an occurrence.
If a single transaction might contain millions of $'s, or loss of life,
then it may not be an acceptable risk.
No publicly quoted company can afford to lose financial transactions, no
matter how small; if the auditors refuse to sign off the accounts, the
whole company is toast.
It would mean write-off's that could be explained. Heck, look at the
billions HP eventually wrote off when they purchased Autonomy.

Actually I heard a rumour that banks will continue to release $'s from
ATM's - even if the back end system is down.

The reasoning was that this event would happen so infrequently and the
amount of $'s they might lose to fraud is so small (in relative terms)
since most clients have daily withdrawal limits, so they would rather
lose a few $'s than upset their Customers i.e. "system unavailable" type
issues.

That's the "rumour" I heard anyway.
Post by Kerry Main
Each business needs to determine how much DR/DT "insurance" they
are
Post by Kerry Main
willing to pay for to address the basic risks, while accepting that if
a significant event does occur, they may be in deep trouble.
Unfortunately, most businesses today are increasingly depending on
IT.
Post by Kerry Main
BA found this out the hard way.
One niggling question about BA's disaster - various commentators were
questioning that if the cause was an upgrade gone wrong (there was a
Tweet which suggested that), why did they do it on a UK holiday weekend
when so many passengers might be affected?
One possible answer to that is that management decided that avoiding a
disruption to Business Class passengers during the week was of a higher
priority. There are different ways to alleviate risks - at one place we
decided that any upgrade taking less than an hour or two happened over a
lunch time, so that in the event of something going wrong, all the
application teams were on hand to diagnose and fix the problem(s). :-)
I had the same thing at a semi chip manufacturing company. We planned
and did online OS/HW upgrades (including rolling reboots of individual
nodes) to a mission critical cluster for 4 months to plan a 1 hour
cluster reboot at noon to finally change the application as well.
Reasoning was the same - 24x7 shop with downtime measured at approx.
$750K/hr, and hence, the best time of day was determined to be noon when
everyone else went to lunch and all the support resources were available
and on site.

Worked perfectly.


Regards,

Kerry Main
Kerry dot main at starkgaming dot com

David Froble
2017-06-18 19:50:37 UTC
Permalink
Raw Message
Post by Kerry Main
-----Original Message-----
Simon Clubley via Info-vax
Sent: June 18, 2017 5:17 AM
Subject: Re: [Info-vax] What is the future role for VMS ?
Post by Kerry Main
Stock exchanges, likely candidates for fault tolerance are evolving.
They now need very high HA - not fault tolerance. Their need for
speed
Post by Kerry Main
has in some cases over ridden the need for fault tolerance. Hence,
it is
Post by Kerry Main
not uncommon to have a stock exchange running in active-passive
mode
Post by Kerry Main
between sites which means that in the event of a 9/11 type incident,
they WILL lose data. The amount of data lost would depend on the
replication timeframe between site updates and when the incident
occurred.
Of course, that worked really well for BA... :-)
Post by Kerry Main
In some cases, you might see a stock exchange say something like "we
gained a lot of speed with our new system!". What they will not say
is
Post by Kerry Main
that they also gave up their RPO=0 requirement.
Is giving up RPO=0 a false economy ?
Simon.
Given the nature of the business, losing data may be an acceptable risk
given the alternative costs to prevent such an occurrence.
If a single transaction might contain millions of $'s, or loss of life,
then it may not be an acceptable risk.
Each business needs to determine how much DR/DT "insurance" they are
willing to pay for to address the basic risks, while accepting that if a
significant event does occur, they may be in deep trouble.
Unfortunately, most businesses today are increasingly depending on IT.
BA found this out the hard way.
Another example - company went bankrupt when they did not follow DR/DT
and security best practices with offsite backups separate from the
cloud.
<http://www.computerworld.com/article/2491008/cloud-security/hacker-puts
--full-redundancy--code-hosting-firm-out-of-business.html>
"CodeSpaces.com shut down after a hacker gained access to its Amazon EC2
account and deleted most data, including backups"
"A code-hosting and project management services provider was forced to
shut down operations indefinitely after a hacker broke into its cloud
infrastructure and deleted customer data, including most of the
company's backups."
With off-site backups, they would have still been impacted, but they
would at least been able to recover.
One thing to keep in mind when a company doesn't cover all the bases. Perhaps
they are looking at:

1) If we don't get faster, we're going out of business, for sure.
2) We MIGHT get hacked, and possibly go out of business, maybe.

Now, which would you pick?

Note, I don't agree that those are the only two choices, and I'm sure I could
come up with alternatives. Still, there might be those who feel they have only
the two choices.
Kerry Main
2017-06-18 20:27:04 UTC
Permalink
Raw Message
-----Original Message-----
David Froble via Info-vax
Sent: June 18, 2017 3:51 PM
Subject: Re: [Info-vax] What is the future role for VMS ?
Post by Kerry Main
-----Original Message-----
Simon Clubley via Info-vax
Sent: June 18, 2017 5:17 AM
Earth.UFP>
Post by Kerry Main
Subject: Re: [Info-vax] What is the future role for VMS ?
Post by Kerry Main
Stock exchanges, likely candidates for fault tolerance are
evolving.
Post by Kerry Main
Post by Kerry Main
They now need very high HA - not fault tolerance. Their need for
speed
Post by Kerry Main
has in some cases over ridden the need for fault tolerance. Hence,
it is
Post by Kerry Main
not uncommon to have a stock exchange running in active-passive
mode
Post by Kerry Main
between sites which means that in the event of a 9/11 type
incident,
Post by Kerry Main
Post by Kerry Main
they WILL lose data. The amount of data lost would depend on the
replication timeframe between site updates and when the incident
occurred.
Of course, that worked really well for BA... :-)
Post by Kerry Main
In some cases, you might see a stock exchange say something like
"we
Post by Kerry Main
Post by Kerry Main
gained a lot of speed with our new system!". What they will not say
is
Post by Kerry Main
that they also gave up their RPO=0 requirement.
Is giving up RPO=0 a false economy ?
Simon.
Given the nature of the business, losing data may be an acceptable risk
given the alternative costs to prevent such an occurrence.
If a single transaction might contain millions of $'s, or loss of life,
then it may not be an acceptable risk.
Each business needs to determine how much DR/DT "insurance" they
are
Post by Kerry Main
willing to pay for to address the basic risks, while accepting that if a
significant event does occur, they may be in deep trouble.
Unfortunately, most businesses today are increasingly depending on
IT.
Post by Kerry Main
BA found this out the hard way.
Another example - company went bankrupt when they did not follow
DR/DT
Post by Kerry Main
and security best practices with offsite backups separate from the
cloud.
<http://www.computerworld.com/article/2491008/cloud-
security/hacker-puts
Post by Kerry Main
--full-redundancy--code-hosting-firm-out-of-business.html>
"CodeSpaces.com shut down after a hacker gained access to its
Amazon EC2
Post by Kerry Main
account and deleted most data, including backups"
"A code-hosting and project management services provider was forced
to
Post by Kerry Main
shut down operations indefinitely after a hacker broke into its cloud
infrastructure and deleted customer data, including most of the
company's backups."
With off-site backups, they would have still been impacted, but they
would at least been able to recover.
One thing to keep in mind when a company doesn't cover all the bases.
Perhaps
1) If we don't get faster, we're going out of business, for sure.
2) We MIGHT get hacked, and possibly go out of business, maybe.
Now, which would you pick?
Note, I don't agree that those are the only two choices, and I'm sure
I
could
come up with alternatives. Still, there might be those who feel they have only
the two choices.
One of the reasons why "public cloud" is cheaper is because they do not
always offer the same services (like offsite backups and/or archiving)
as traditional IT shops which are required to properly support the
company business.

When asked about backups, sadly, all to often the Cloud provider
response is either "we have a number of replication options" (as if
replication was the same as backup) or "you can continue doing backups
on your own" (which means additional licensing $'s, additional storage,
additional bandwidth, extra staff etc.) that is not included in the
upfront pricing which is often used to justify the move to cloud.

In addition to security concerns, its why a number of early adopter
companies in both the US and Europe are now moving their "Public Cloud"
business back in house or as it is now called "back on-premises".

June 16, 2017:
<http://www.eweek.com/storage/pure-storage-goes-big-with-flash-object-st
orage-new-ai-engine>
"Despite strong overall indicators of public cloud growth, a significant
number of companies that ran workloads in public cloud environments have
actually moved some or all of those workloads back on-premises (43
percent of businesses in North America have done so, according to the
survey). In EMEA, 65 percent say they have reduced use of public cloud
in the last 12 months because of security concerns."

Regards,

Kerry Main
Kerry dot main at starkgaming dot com
Stephen Hoffman
2017-06-12 18:34:21 UTC
Permalink
Raw Message
Post by IanD
I see a real differentiator as being a system where a process can be
kept infinitely alive and migrated anywhere
OpenVMS tried parts of that many years ago, and ran afoul of the
complexity of the process and system context, both around I/O and
shared I/O contexts, and around clustering and network state.
Reestablishing the channel context with the other applications that
might have the file open shared, re-opening I/O channels to other hosts
or other cluster members, etc.

This with checkpoint-restart investigations, and the so-called fastboot
(fboot, et al) support work from VAX workstation days.

This whole area is also tied into online backups of active applications
and application data. Any sort of migration is a close relative of —
if not a superset of — getting a good consistent backup.

Many of the virtual machine failover demos have involved shorter and
shorter snapshots of the running system, and then failing over and
restarting that whole system, and I'm not at all sure how well that'll
work with the sort of complex context that OpenVMS apps drag around,
particularly in clusters and networked applications, and device state
with FC SAN storage or NAS storage that OpenVMS involves. This
reminds me of the same problems that lurk behind storage-level
replication — without host coordination and assistance, that's very
tough to get right.)
Post by IanD
If you computing a huge workload that run for say 2 days and then the
system dies, what good is a cluster?
That's been possible for many years, either with intermediate storage
stored in a transactional database, or with an app coded to use
transactions or analogous.

This all requires both operating system modifications and custom coding
added into the application.
Post by IanD
I want to be able to shadow my process n x's for as much redundancy as I wish
That's also been possible with app mods, but — if you want to buy tools
and APIs to add more redundancy — it's closer to what NSK offers.

Whether enough OpenVMS customers will accept the need for app
modifications, for transactional databases, or configuring and running
something akin to Hadoop? Donno. Hive atop Hadoop might interest some
folks, but those folks are probably already running SQL databases on
OpenVMS. Folks that are used to the traditional OpenVMS batch
environment might find an alternative akin to HTcondor more familiar
and more palatable, too.

These days, the performance of the old VAX fastboot support — the
now-deprecated FBOOT parameter et al — is positively glacial when
compared with the typical cached boot performance found on macOS or
other boxes, too. Times and techniques change.

TL;DR: If any sort of migration is to be tried again, it'll almost
certainly require application modifications to communicate state with
the operating system (cluster distributed scheduler, failover support,
online backup, etc).
--
Pure Personal Opinion | HoffmanLabs LLC
Arne Vajhøj
2017-06-12 18:41:59 UTC
Permalink
Raw Message
Post by Stephen Hoffman
Post by IanD
I see a real differentiator as being a system where a process can be
kept infinitely alive and migrated anywhere
OpenVMS tried parts of that many years ago, and ran afoul of the
complexity of the process and system context, both around I/O and shared
I/O contexts, and around clustering and network state. Reestablishing
the channel context with the other applications that might have the file
open shared, re-opening I/O channels to other hosts or other cluster
members, etc.
"Nothing in the world is worth having or worth doing unless it means
effort, pain, difficulty"

:-)

Arne
Stephen Hoffman
2017-06-12 19:33:14 UTC
Permalink
Raw Message
Post by Arne Vajhøj
Post by Stephen Hoffman
Post by IanD
I see a real differentiator as being a system where a process can be
kept infinitely alive and migrated anywhere
OpenVMS tried parts of that many years ago, and ran afoul of the
complexity of the process and system context, both around I/O and
shared I/O contexts, and around clustering and network state.
Reestablishing the channel context with the other applications that
might have the file open shared, re-opening I/O channels to other hosts
or other cluster members, etc.
"Nothing in the world is worth having or worth doing unless it means
effort, pain, difficulty"
:-)
Arne
So... VSI will be porting OpenVMS over to an L4 kernel? As for
product development... "Do what you can, with what you have, where
you are." TR quotes and jokes aside... More than a few development
projects get axed when the expected effort is higher than the expected
revenue produced, or when the length of the expected development effort
is longer than the resulting product or service is expected t be
profitable in the market, or when the risk to the business is
considered too high. The projects can and have to do what they can,
within the available schedule and staffing and budget.
--
Pure Personal Opinion | HoffmanLabs LLC
Arne Vajhøj
2017-06-12 20:15:51 UTC
Permalink
Raw Message
Post by Stephen Hoffman
Post by Arne Vajhøj
Post by Stephen Hoffman
Post by IanD
I see a real differentiator as being a system where a process can be
kept infinitely alive and migrated anywhere
OpenVMS tried parts of that many years ago, and ran afoul of the
complexity of the process and system context, both around I/O and
shared I/O contexts, and around clustering and network state.
Reestablishing the channel context with the other applications that
might have the file open shared, re-opening I/O channels to other
hosts or other cluster members, etc.
"Nothing in the world is worth having or worth doing unless it means
effort, pain, difficulty"
:-)
So... VSI will be porting OpenVMS over to an L4 kernel? As for
product development... "Do what you can, with what you have, where you
are." TR quotes and jokes aside... More than a few development
projects get axed when the expected effort is higher than the expected
revenue produced, or when the length of the expected development effort
is longer than the resulting product or service is expected t be
profitable in the market, or when the risk to the business is considered
too high. The projects can and have to do what they can, within the
available schedule and staffing and budget.
Someone need to estimate the cost and the potential revenue.

But I will recommend not being too fast to reject ideas.

Especially not if the ideas are targeting a market that
in general is relevant for VMS.

Arne
Stephen Hoffman
2017-06-12 21:14:21 UTC
Permalink
Raw Message
Post by Arne Vajhøj
Someone need to estimate the cost and the potential revenue.
But I will recommend not being too fast to reject ideas.
Especially not if the ideas are targeting a market that in general is
relevant for VMS.
Sorry, no TR quote here. Just a code-slog.

Anybody that wants to can implement app restarts on OpenVMS. Right
now. For many environments, the effort involved is fussy and arcane.
There are many easy mistakes possible, too. But there's nothing here
that's particularly unique or unusual. Simplest approach starts with
the batch restart support that few use. SET RESTART_VALUE,
BATCH$RESTART, et al. More involved approaches can combine clustering
and shadowing, DEdtm distributed transaction manager, maybe RMS
journaling, or uses message queuing, or one of the previously-cited
frameworks. Inevitably distributed authentication and encryption.
Not that I'd consider any of these to be particularly modern
implementations for the developers to deal with.

Many developers ignore some or all of these available options.
Whether intentionally, or through omission or ignorance of the features
and capabilities, or they've been dissuaded by the license prices
involved?

From what's involved here in the layered products and in the OpenVMS
operating system, the integration, incorporation and further
abstraction of these existing features (and maybe adding something
equivalent to DECscheduler or HTcondor or another scheduler into the
mix) requires a decent investment in design and implementation in the
operating system, and it requires application modifications —
communications of state between the system and the apps, as well as
managing details such as distributed authentication and encryption, and
state change notifications — if this whole environment is to become
somewhat more transparent to the app developers. Some folks point to
the class scheduler, but in this context control groups (cgroups)
support provides rather more. For various folks, using a database for
this general purpose — online backups, recovery from failures,
restarting from some intermediate app state — works well enough.

Creating a simple and clear and easy-to-use design that integrate all
of these existing APIs and products — and whatever new giblets might be
added — is an order of magnitude harder, too.

Yet harder? Implementing something that's transparent to existing
apps is somewhere between a huge effort, an intractable effort, or
simply impossible. Not with all the app state baggage and network and
cluster baggage involved underneath all but the most trivial of OpenVMS
apps.

VSI mentions some semi-related research in this direction in their
roadmap (containers, et al), but there's still a lot of code to write
or port, and test and document and...

Just getting the whole development and APIs better documented —
including dragging the security manual forward by a quarter-century or
so — would be a start.
--
Pure Personal Opinion | HoffmanLabs LLC
Kerry Main
2017-06-17 20:18:28 UTC
Permalink
Raw Message
-----Original Message-----
Stephen Hoffman via Info-vax
Sent: June 12, 2017 5:14 PM
Subject: Re: [Info-vax] What is the future role for VMS ?
Post by Arne Vajhøj
Someone need to estimate the cost and the potential revenue.
But I will recommend not being too fast to reject ideas.
Especially not if the ideas are targeting a market that in general is
relevant for VMS.
[snip]

An interesting view from 2067 looking back .. I wonder where OpenVMS will be in this next gen view of the world?

< https://medium.com/iotforall/the-year-is-2067-a95a786994c5>
"As he started to read the fifty-year-old article, the first line literally made him guffaw. The author, Jared Porcenaluk, wrote “The Internet of Things is shaping up to change our world in dramatic ways.”

“You don’t say,” thought John."

😊

Regards,

Kerry Main
Kerry dot main at starkgaming dot com
u***@gmail.com
2017-06-07 03:54:54 UTC
Permalink
Raw Message
Post by Simon Clubley
Since it's come up once again, it's time to ask people outright
what market segments you see as being a possible future for VMS.
Is x86-64 VMS a server OS for both new and existing users ?
Is x86-64 VMS a server OS for existing users only ?
Is x86-64 VMS a desktop OS (at least for some people) ?
Is x86-64 VMS some kind of embedded/process control OS ?
I think the only possible future for VMS for the next 5+ years is
going to be for existing users looking to keep VMS applications
running within their organisation.
Past that, you are going to need some unique application or feature
VMS can offer which would be unique enough that it would persuade
new people to try it and VMS.
However, I do not currently see what that can be.
Do you see such a possible unique application or feature ?
Simon.
--
Microsoft: Bringing you 1980s technology to a 21st century world
I see one unique feature, security ;)
David Froble
2017-06-07 04:22:46 UTC
Permalink
Raw Message
Post by u***@gmail.com
Post by Simon Clubley
Since it's come up once again, it's time to ask people outright
what market segments you see as being a possible future for VMS.
Is x86-64 VMS a server OS for both new and existing users ?
Is x86-64 VMS a server OS for existing users only ?
Is x86-64 VMS a desktop OS (at least for some people) ?
Is x86-64 VMS some kind of embedded/process control OS ?
I think the only possible future for VMS for the next 5+ years is
going to be for existing users looking to keep VMS applications
running within their organisation.
Past that, you are going to need some unique application or feature
VMS can offer which would be unique enough that it would persuade
new people to try it and VMS.
However, I do not currently see what that can be.
Do you see such a possible unique application or feature ?
Simon.
--
Microsoft: Bringing you 1980s technology to a 21st century world
I see one unique feature, security ;)
If you're so sure of that, maybe you'd post the IP address of your VMS system,
and set up a non-prived guest user account (posting that username and password)
for some people who may have idle time on their hands?

Gauntlet thrown down, will it be picked up?
Stephen Hoffman
2017-06-07 16:01:41 UTC
Permalink
Raw Message
Post by u***@gmail.com
I see one unique feature, security ;)
Even assuming facts not in evidence, security is only a selling point
rather further down the purchase requirements list, after the platform
supports the necessary applications, and after the platform is found to
be affordable.
--
Pure Personal Opinion | HoffmanLabs LLC
u***@gmail.com
2017-06-07 03:56:35 UTC
Permalink
Raw Message
Post by Simon Clubley
Since it's come up once again, it's time to ask people outright
what market segments you see as being a possible future for VMS.
Is x86-64 VMS a server OS for both new and existing users ?
Is x86-64 VMS a server OS for existing users only ?
Is x86-64 VMS a desktop OS (at least for some people) ?
Is x86-64 VMS some kind of embedded/process control OS ?
I think the only possible future for VMS for the next 5+ years is
going to be for existing users looking to keep VMS applications
running within their organisation.
Past that, you are going to need some unique application or feature
VMS can offer which would be unique enough that it would persuade
new people to try it and VMS.
However, I do not currently see what that can be.
Do you see such a possible unique application or feature ?
Simon.
--
Microsoft: Bringing you 1980s technology to a 21st century world
and before you chime in with the linux has security bit, just recall
vms was written to correct unix deficencies. Linux is a poor mans unix.
I think that says it all.
David Froble
2017-06-07 04:24:37 UTC
Permalink
Raw Message
Post by u***@gmail.com
Post by Simon Clubley
Since it's come up once again, it's time to ask people outright
what market segments you see as being a possible future for VMS.
Is x86-64 VMS a server OS for both new and existing users ?
Is x86-64 VMS a server OS for existing users only ?
Is x86-64 VMS a desktop OS (at least for some people) ?
Is x86-64 VMS some kind of embedded/process control OS ?
I think the only possible future for VMS for the next 5+ years is
going to be for existing users looking to keep VMS applications
running within their organisation.
Past that, you are going to need some unique application or feature
VMS can offer which would be unique enough that it would persuade
new people to try it and VMS.
However, I do not currently see what that can be.
Do you see such a possible unique application or feature ?
Simon.
--
Microsoft: Bringing you 1980s technology to a 21st century world
and before you chime in with the linux has security bit, just recall
vms was written to correct unix deficencies. Linux is a poor mans unix.
I think that says it all.
Me thinks you've been smoking some bad stuff. The design and implementation of
VMS had nothing to do with *ix.
Simon Clubley
2017-06-07 17:47:33 UTC
Permalink
Raw Message
Post by David Froble
Post by u***@gmail.com
and before you chime in with the linux has security bit, just recall
vms was written to correct unix deficencies. Linux is a poor mans unix.
I think that says it all.
Me thinks you've been smoking some bad stuff. The design and implementation of
VMS had nothing to do with *ix.
In addition, thanks to investment, Linux and other Unix environments
have had their security environments enhanced over the years while
VMS has remained stagnant.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world
IanD
2017-06-07 06:42:59 UTC
Permalink
Raw Message
I have indirectly asked this question before in other subtle ways

While these are not exhaustive market segments, I'm hard pressed to think where VMS could dominate

- Embedded (linux)
- Mobile (linux / android / apple)
- Virtualisation (linux again)
- Desktop (windows)
- Enterprise applications (linux / windows)
- HPC (linux)
- Security (??? this is an area with big returns and high risk if you fail)
- Clustering (linux again. It doesn't matter that they usurped the cluster name)
- Development (windows, linux)
- Education (linux again)
- IoT (linux, windows trying to gain traction)
- Opensource (I know its not a market segment but it's a driver, linux)
- Blockchain (linix, yet again)

As much as I love VMS, I am hard pressed to think of an area where it will compete and dominate or at least grow

yeah yeah yeah, don't be negative, don't poo-poo it before it's even got onto x86. I'm not, I just needing help in seeing a clear market segment where it can lever itself into and grow - enlighten me

Now for some tongue in cheek, i know one area...

- Nostalgic computing :-) (take this will the good humor it's intended to be in)
Jan-Erik Soderholm
2017-06-07 08:42:22 UTC
Permalink
Raw Message
Post by IanD
I have indirectly asked this question before in other subtle ways
While these are not exhaustive market segments, I'm hard pressed to think where VMS could dominate
- Embedded (linux)
- Mobile (linux / android / apple)
- Virtualisation (linux again)
- Desktop (windows)
- Enterprise applications (linux / windows)
- HPC (linux)
- Security (??? this is an area with big returns and high risk if you fail)
- Clustering (linux again. It doesn't matter that they usurped the cluster name)
- Development (windows, linux)
- Education (linux again)
- IoT (linux, windows trying to gain traction)
- Opensource (I know its not a market segment but it's a driver, linux)
- Blockchain (linix, yet again)
As much as I love VMS, I am hard pressed to think of an area where it will compete and dominate or at least grow
yeah yeah yeah, don't be negative, don't poo-poo it before it's even got onto x86. I'm not, I just needing help in seeing a clear market segment where it can lever itself into and grow - enlighten me
Now for some tongue in cheek, i know one area...
- Nostalgic computing :-) (take this will the good humor it's intended to be in)
Industrial and/or Production support?
j***@yahoo.co.uk
2017-06-07 09:11:20 UTC
Permalink
Raw Message
Post by Jan-Erik Soderholm
Post by IanD
I have indirectly asked this question before in other subtle ways
While these are not exhaustive market segments, I'm hard pressed to think where VMS could dominate
- Embedded (linux)
- Mobile (linux / android / apple)
- Virtualisation (linux again)
- Desktop (windows)
- Enterprise applications (linux / windows)
- HPC (linux)
- Security (??? this is an area with big returns and high risk if you fail)
- Clustering (linux again. It doesn't matter that they usurped the cluster name)
- Development (windows, linux)
- Education (linux again)
- IoT (linux, windows trying to gain traction)
- Opensource (I know its not a market segment but it's a driver, linux)
- Blockchain (linix, yet again)
As much as I love VMS, I am hard pressed to think of an area where it will compete and dominate or at least grow
yeah yeah yeah, don't be negative, don't poo-poo it before it's even got onto x86. I'm not, I just needing help in seeing a clear market segment where it can lever itself into and grow - enlighten me
Now for some tongue in cheek, i know one area...
- Nostalgic computing :-) (take this will the good humor it's intended to be in)
Industrial and/or Production support?
Now we might be getting somewhere (depending on what
exactly you mean by industrial/production support).

High value automation in general, which might include
stuff that (encouraged by HP) has moved (uncomfortably)
off VMS to places where it didn't really belong.

Presumably this might include related capabilities
outside factories too, e.g. parts of command and
control systems, emergency services despatch systems,
traffic control systems, and so on.

Call it SCADA, if you will, as a starting point. But
simple low value low risk SCADA/automation may be
someone else's opportunity for now.

It may not be a shiny high visibility huge volume
market but that doesn't necessarily mean it isn't a
financially worthwhile market for someone.

ps
VMS in the traffic management market is
Variable Matrrix Sign
Jan-Erik Soderholm
2017-06-07 10:15:26 UTC
Permalink
Raw Message
Post by j***@yahoo.co.uk
Post by Jan-Erik Soderholm
Post by IanD
I have indirectly asked this question before in other subtle ways
While these are not exhaustive market segments, I'm hard pressed to think where VMS could dominate
- Embedded (linux)
- Mobile (linux / android / apple)
- Virtualisation (linux again)
- Desktop (windows)
- Enterprise applications (linux / windows)
- HPC (linux)
- Security (??? this is an area with big returns and high risk if you fail)
- Clustering (linux again. It doesn't matter that they usurped the cluster name)
- Development (windows, linux)
- Education (linux again)
- IoT (linux, windows trying to gain traction)
- Opensource (I know its not a market segment but it's a driver, linux)
- Blockchain (linix, yet again)
As much as I love VMS, I am hard pressed to think of an area where it will compete and dominate or at least grow
yeah yeah yeah, don't be negative, don't poo-poo it before it's even got onto x86. I'm not, I just needing help in seeing a clear market segment where it can lever itself into and grow - enlighten me
Now for some tongue in cheek, i know one area...
- Nostalgic computing :-) (take this will the good humor it's intended to be in)
Industrial and/or Production support?
Now we might be getting somewhere (depending on what
exactly you mean by industrial/production support).
Probably the area that is also covered by what is called MES.

https://en.wikipedia.org/wiki/Manufacturing_execution_system

Probably focused at level 3 of the "Purdue Reference Model 95”.

https://en.wikipedia.org/wiki/Purdue_Enterprise_Reference_Architecture

SCADA (mentioned below) are usually on a lower level (2).

MES systems is the "glue" between ERP and SCADA (and lower).
Post by j***@yahoo.co.uk
High value automation in general, which might include
stuff that (encouraged by HP) has moved (uncomfortably)
off VMS to places where it didn't really belong.
Presumably this might include related capabilities
outside factories too, e.g. parts of command and
control systems, emergency services despatch systems,
traffic control systems, and so on.
Call it SCADA, if you will, as a starting point. But
simple low value low risk SCADA/automation may be
someone else's opportunity for now.
It may not be a shiny high visibility huge volume
market but that doesn't necessarily mean it isn't a
financially worthwhile market for someone.
ps
VMS in the traffic management market is
Variable Matrrix Sign
Or: http://www.vagmaskinservice.se/

"Road Machinery Service", in English... :-)
j***@yahoo.co.uk
2017-06-07 12:52:08 UTC
Permalink
Raw Message
Post by Jan-Erik Soderholm
Post by j***@yahoo.co.uk
Post by Jan-Erik Soderholm
Post by IanD
I have indirectly asked this question before in other subtle ways
While these are not exhaustive market segments, I'm hard pressed to think where VMS could dominate
- Embedded (linux)
- Mobile (linux / android / apple)
- Virtualisation (linux again)
- Desktop (windows)
- Enterprise applications (linux / windows)
- HPC (linux)
- Security (??? this is an area with big returns and high risk if you fail)
- Clustering (linux again. It doesn't matter that they usurped the cluster name)
- Development (windows, linux)
- Education (linux again)
- IoT (linux, windows trying to gain traction)
- Opensource (I know its not a market segment but it's a driver, linux)
- Blockchain (linix, yet again)
As much as I love VMS, I am hard pressed to think of an area where it will compete and dominate or at least grow
yeah yeah yeah, don't be negative, don't poo-poo it before it's even got onto x86. I'm not, I just needing help in seeing a clear market segment where it can lever itself into and grow - enlighten me
Now for some tongue in cheek, i know one area...
- Nostalgic computing :-) (take this will the good humor it's intended to be in)
Industrial and/or Production support?
Now we might be getting somewhere (depending on what
exactly you mean by industrial/production support).
Probably the area that is also covered by what is called MES.
https://en.wikipedia.org/wiki/Manufacturing_execution_system
Probably focused at level 3 of the "Purdue Reference Model 95”.
https://en.wikipedia.org/wiki/Purdue_Enterprise_Reference_Architecture
SCADA (mentioned below) are usually on a lower level (2).
MES systems is the "glue" between ERP and SCADA (and lower).
Post by j***@yahoo.co.uk
High value automation in general, which might include
stuff that (encouraged by HP) has moved (uncomfortably)
off VMS to places where it didn't really belong.
Presumably this might include related capabilities
outside factories too, e.g. parts of command and
control systems, emergency services despatch systems,
traffic control systems, and so on.
Call it SCADA, if you will, as a starting point. But
simple low value low risk SCADA/automation may be
someone else's opportunity for now.
It may not be a shiny high visibility huge volume
market but that doesn't necessarily mean it isn't a
financially worthwhile market for someone.
ps
VMS in the traffic management market is
Variable Matrrix Sign
Or: http://www.vagmaskinservice.se/
"Road Machinery Service", in English... :-)
MES could be part of it, yes. People who think
in terms of MES may well also be thinking in terms
of other well-established three-letter packages e.g.
from SAP and friends, so may not be immediate
candidates for VSIVMS. Stranger things have
happened in the past though.

Some of these systems (not just MES) tend to need
"sufficient" performance rather than massive
performance, whereas downtime on (e.g.) a high value
manufacturing line might be very expensive and might
justify something a little non-mainstream that doesn't
require downtime (planned or unplanned) every few weeks.

My crystal ball is in an extended period of downtime
right now, maybe I'll come back in a couple of years
when things might be clearer.
Kerry Main
2017-06-10 14:51:47 UTC
Permalink
Raw Message
-----Original Message-----
johnwallace4--- via Info-vax
Sent: June 7, 2017 8:52 AM
Subject: Re: [Info-vax] What is the future role for VMS ?
Post by Jan-Erik Soderholm
On Wednesday, 7 June 2017 09:42:24 UTC+1, Jan-Erik Soderholm
Post by Jan-Erik Soderholm
Post by IanD
I have indirectly asked this question before in other subtle ways
While these are not exhaustive market segments, I'm hard
pressed to think where VMS could dominate
Post by Jan-Erik Soderholm
Post by Jan-Erik Soderholm
Post by IanD
- Embedded (linux)
- Mobile (linux / android / apple)
- Virtualisation (linux again)
- Desktop (windows)
- Enterprise applications (linux / windows)
- HPC (linux)
- Security (??? this is an area with big returns and high risk if you
fail)
Post by Jan-Erik Soderholm
Post by Jan-Erik Soderholm
Post by IanD
- Clustering (linux again. It doesn't matter that they usurped the
cluster name)
Post by Jan-Erik Soderholm
Post by Jan-Erik Soderholm
Post by IanD
- Development (windows, linux)
- Education (linux again)
- IoT (linux, windows trying to gain traction)
- Opensource (I know its not a market segment but it's a driver,
linux)
Post by Jan-Erik Soderholm
Post by Jan-Erik Soderholm
Post by IanD
- Blockchain (linix, yet again)
As much as I love VMS, I am hard pressed to think of an area
where it will compete and dominate or at least grow
Post by Jan-Erik Soderholm
Post by Jan-Erik Soderholm
Post by IanD
yeah yeah yeah, don't be negative, don't poo-poo it before it's
even got onto x86. I'm not, I just needing help in seeing a clear market
segment where it can lever itself into and grow - enlighten me
Post by Jan-Erik Soderholm
Post by Jan-Erik Soderholm
Post by IanD
Now for some tongue in cheek, i know one area...
- Nostalgic computing :-) (take this will the good humor it's
intended to be in)
Post by Jan-Erik Soderholm
Post by Jan-Erik Soderholm
Industrial and/or Production support?
Now we might be getting somewhere (depending on what
exactly you mean by industrial/production support).
Probably the area that is also covered by what is called MES.
https://en.wikipedia.org/wiki/Manufacturing_execution_system
Probably focused at level 3 of the "Purdue Reference Model 95”.
https://en.wikipedia.org/wiki/Purdue_Enterprise_Reference_Architectu
re
Post by Jan-Erik Soderholm
SCADA (mentioned below) are usually on a lower level (2).
MES systems is the "glue" between ERP and SCADA (and lower).
High value automation in general, which might include
stuff that (encouraged by HP) has moved (uncomfortably)
off VMS to places where it didn't really belong.
Presumably this might include related capabilities
outside factories too, e.g. parts of command and
control systems, emergency services despatch systems,
traffic control systems, and so on.
Call it SCADA, if you will, as a starting point. But
simple low value low risk SCADA/automation may be
someone else's opportunity for now.
It may not be a shiny high visibility huge volume
market but that doesn't necessarily mean it isn't a
financially worthwhile market for someone.
ps
VMS in the traffic management market is
Variable Matrrix Sign
Or: http://www.vagmaskinservice.se/
"Road Machinery Service", in English... :-)
MES could be part of it, yes. People who think
in terms of MES may well also be thinking in terms
of other well-established three-letter packages e.g.
from SAP and friends, so may not be immediate
candidates for VSIVMS. Stranger things have
happened in the past though.
[snip]

Keep in mind that SAP used to run on OpenVMS, so I agree - never say never. I still have a "SAP Install on OpenVMS" manual somewhere (think it was Alpha?).

Perhaps at a future point, it might be time to dust off the SAP on OpenVMS code?

Perhaps having an integrated multi-site cluster that looks like a single system to developers and end users might again become popular once OpenVMS is on X86-64.

Note - a few years back, a big push by SAP was to put the App server and the DB server on to the same OS as a means to consolidate OS's and reduce App-DB latency. Remember that SAP roots are in centralized systems with big batch job queues.

For OpenVMS environments, other than the traditional best practices associated with centralized capacity planning and testing (remember these often forgotten concepts?) this would not be an issue at all (culturally or technically).

For Windows/Linux environments, these types of App stacking directions are going to be very challenging. The cultural challenges alone are huge i.e. "No way I would ever put more than one Bus App on the same server OS ...".

Especially with the DC world future moving to much more centralized models than the distributed models so popular in the last 20 years.

Book: SAP Hardware Solutions: Servers, Storage, and Networks for MySAP.com
< http://bit.ly/2seDF8y>
< https://books.google.ca/books?id=IEtJmO0FuawC&printsec=frontcover#v=onepage&q&f=false>

Other big application platforms are looking at similar consolidation initiatives.

Future compute models are changing drastically to accommodate the rapidly changing technologies, so imho, there are all sorts of opportunities for OpenVMS in the X86-64 world.


Regards,

Kerry Main
Kerry dot main at starkgaming dot com
IanD
2017-06-09 05:42:54 UTC
Permalink
Raw Message
<snip>
Post by j***@yahoo.co.uk
ps
VMS in the traffic management market is
Variable Matrrix Sign
The SCATS traffic system, sold worldwide I believe used to be on PDP from what I was told

Then to Vax?

A couple of years ago I was waiting to go into a kids concert and I overheard someone talking about SCATS. i struck up on conversation and asked about it from the aspect of 'didn't that used to run on VMS?'

The person then went on to explain that yes, but it was ported to windows server now and ran very well and any notion of it going back to VMS would 'never be entertained'

Another area for VMS is Steel mills - isn't VMS still running a lot of the mills out there? I have also heard its in some nuclear power control systems? I just heard this, I'd love to see proof though
clair.grant@vmssoftware.com
2017-06-09 11:35:20 UTC
Permalink
Raw Message
Post by IanD
<snip>
Post by j***@yahoo.co.uk
ps
VMS in the traffic management market is
Variable Matrrix Sign
The SCATS traffic system, sold worldwide I believe used to be on PDP from what I was told
Then to Vax?
A couple of years ago I was waiting to go into a kids concert and I overheard someone talking about SCATS. i struck up on conversation and asked about it from the aspect of 'didn't that used to run on VMS?'
The person then went on to explain that yes, but it was ported to windows server now and ran very well and any notion of it going back to VMS would 'never be entertained'
Another area for VMS is Steel mills - isn't VMS still running a lot of the mills out there? I have also heard its in some nuclear power control systems? I just heard this, I'd love to see proof though
We are in contact with people running VMS in steel mills and nuclear power plants.
Jan-Erik Soderholm
2017-06-09 12:49:15 UTC
Permalink
Raw Message
Post by ***@vmssoftware.com
Post by IanD
<snip>
Post by j***@yahoo.co.uk
ps
VMS in the traffic management market is
Variable Matrrix Sign
The SCATS traffic system, sold worldwide I believe used to be on PDP from what I was told
Then to Vax?
A couple of years ago I was waiting to go into a kids concert and I overheard someone talking about SCATS. i struck up on conversation and asked about it from the aspect of 'didn't that used to run on VMS?'
The person then went on to explain that yes, but it was ported to windows server now and ran very well and any notion of it going back to VMS would 'never be entertained'
Another area for VMS is Steel mills - isn't VMS still running a lot of the mills out there? I have also heard its in some nuclear power control systems? I just heard this, I'd love to see proof though
We are in contact with people running VMS in steel mills and nuclear power plants.
There is one company in Sweden running a lot of VMS in there rolling mill
plant. The Swedish VMS SIG in HP-Connect had one meeting located there
some yers ago. Some 10s of systems in the prod environmen.

At lest France has some nuclear power plants running VMS. That has been
mentioned here before...
David Froble
2017-06-07 14:05:21 UTC
Permalink
Raw Message
Post by IanD
I have indirectly asked this question before in other subtle ways
While these are not exhaustive market segments, I'm hard pressed to think where VMS could dominate
- Embedded (linux)
- Mobile (linux / android / apple)
- Virtualisation (linux again)
- Desktop (windows)
- Enterprise applications (linux / windows)
- HPC (linux)
- Security (??? this is an area with big returns and high risk if you fail)
- Clustering (linux again. It doesn't matter that they usurped the cluster name)
- Development (windows, linux)
- Education (linux again)
- IoT (linux, windows trying to gain traction)
- Opensource (I know its not a market segment but it's a driver, linux)
- Blockchain (linix, yet again)
As much as I love VMS, I am hard pressed to think of an area where it will compete and dominate or at least grow
yeah yeah yeah, don't be negative, don't poo-poo it before it's even got onto x86. I'm not, I just needing help in seeing a clear market segment where it can lever itself into and grow - enlighten me
Now for some tongue in cheek, i know one area...
- Nostalgic computing :-) (take this will the good humor it's intended to be in)
In some of the categories you mention, there is limited to no revenue available.
I'd also mention that you've most likely mentioned just a small part of
possible uses.

But, for a current example, how about airline reservations?

:-)

Ok, but not even VMS is going to help if there is a poor design, or incapable
operators, and such.

Enterprise applications is sort of a vague definition. Regardless, VMS is
useful in many of the categories that might fit in such a category.

Production control
Distribution
Accounting
Planning
Warehouse management
Purchasing

And many more ....
Simon Clubley
2017-06-07 18:17:13 UTC
Permalink
Raw Message
Post by David Froble
But, for a current example, how about airline reservations?
:-)
I think IBM might have a thing or two to say about that... :-)
Post by David Froble
Ok, but not even VMS is going to help if there is a poor design, or incapable
operators, and such.
On one level I strongly agree (especially about the poor design), but at
another level, surely a well designed system can help protect against
various levels of stupidity.

For example, BA would not have experienced the disaster it just has if
it had invested resources to keep the backup data centres in lock-step
with the primary data centre. I've read Scott's comments about higher
transaction latency but surely for something as critical as BA's IT
systems, then investing the resources to overcome that is something
that's justified ?

BTW, I also want to say thank you to everyone who has commented in this
thread so far.

Simon.
--
Simon Clubley, ***@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980s technology to a 21st century world
Arne Vajhøj
2017-06-12 17:10:14 UTC
Permalink
Raw Message
Post by IanD
I have indirectly asked this question before in other subtle ways
While these are not exhaustive market segments, I'm hard pressed to think
where VMS could dominate
- Embedded (linux)
- Mobile (linux / android / apple)
- Virtualisation (linux again)
- Desktop (windows)
- Enterprise applications (linux / windows)
- HPC (linux)
- Security (??? this is an area with big returns and high risk if you fail)
- Clustering (linux again. It doesn't matter that they usurped the cluster name)
- Development (windows, linux)
- Education (linux again)
- IoT (linux, windows trying to gain traction)
- Opensource (I know its not a market segment but it's a driver, linux)
- Blockchain (linix, yet again)
As much as I love VMS, I am hard pressed to think of an area where it will
compete and dominate or at least grow
I think it will be extremely difficult to get Linux users
to switch to VMS while keeping VSI making money.

I don't think VSI can make money selling VMS at RHEL price. And
definitely not at CentOS price.

But there are other OS'es out there as well.

IBM iSeries, commercial Unixes (AIX, Solaris, HP-UX),
highend Windows.

These customers are used to pay for stuff.

Highend Windows is not cheap.

And with VMS being ported to x86-64 and VSI fully dedicated
to VMS, then VMS may have a more certain future than the
commercial Unixes.

Arne
Loading...