On 17/03/2020 09:17, Mark Tinka wrote:
On 16/Mar/20 16:40, Mike Bolitho wrote:
I think there's a bit of a misunderstanding of what I'm trying to say
here. We have dual private lines from two Tier I providers. These
interconnect all major hospitals and our data centers. We also have a
third metro connection that connects things regionally. We have DIA
on top of that. I think people are vastly underestimating just how
much $aaS there is within the medical field. TeleDoc, translation
services, remote radiologists, the way prescriptions get filled, how
staffing works, third party providers basically hoteling within our
facilities, critical staff VPNed in because the government has locked
things down, etc. Then there's things that we don't use but I'm sure
other providers do, GoToMeeting, O365, VaaS, etc.There's no practical
way to engineer your WAN to facilitate dozens of connections to these
services.
This extends beyond just hospitals as well. Fire departments, police
departments, water treatment etc. Regardless of whether or not those
entities planned well(I think we did), the government should and will
step in if critical services are degraded. And for what it's worth,
Stephen, I know how things are built within the ISP world. I spent
four years there. That doesn't change the fact that we're possibly
heading into uncharted waters when it comes to utilization and the
impactthat will have on $aaS products that are interwoven into every
single vertical, including entities that fall under TSP, critical
national security and emergency preparedness functions, including
those areas related to safety, maintenance of law and order, and
public health.It's easy for all you guys to sit here and armchair
quarterback other people's planning but when things really start to
degrade, all bets are off.If you don't believe that, just look at the
news. States are literally shutting down private businesses
(restaurants, bars, night clubs, private schools) and banning people
from associating in groups of larger than 50.
The Internet has infiltrated every industry, every business, and every
business model.
While it's a great way to connect a lot of people and things at scale
for the lowest cost possible, there are some industries that still
require a certain caliber of reliability that the public Internet may
not be best suited to provide.
In your case, I am not sure I have an answer for you, unfortunately.
The public Internet is what it is, mostly best-effort. Your
applications and use-cases certainly deserve better than that. I'm not
sure how to achieve that as your industry shoves more and more
activity into the public Internet domain, for one reason or another.
Mark.
In theory best-effort Internet is seen as only part of a broader
Internet model including open peering and so on. The idea for open
Internet is it offers a form of digital herd immunity (to coin a current
phrase being misused by UK Government circles in recent days) that
offers a level of shared redundancy of spare capacity so that issues can
be taken out of route until fixed but the edge still maintains high
quality connectivity. In one sense the Internet model provides an
informal community insurance across the provider / access sector.
Although of course the legacy telco regulated protected infrastructure
has remained a nub of resistance to open anything.
Some short term financial optimisations between networks may turn out to
be counter productive across time and "events". Which begs a question
whether the winner takes all model that has emerged can live with a
plural supply chain of network infrastructures.
I suspect the concentration over recent years has created greater
fragility for all of us judging comments in this thread and elsewhere.
Can we survive covid 19 and maintain selfish networks over open ones?
C