What are virtual routers and how can they lead to virtual data centers?
So what can you do with virtual router technology? The number one application,
according to enterprises, is virtual networking, especially SD-WAN. All
virtual-network technologies build an overlay network that has its own on- and
off-ramp elements, which are really access routers. While many vendors offer
this technology as appliances, most will also provide virtual routers for
hosting on servers. That may make sense in the data center, where there are
already racks of servers installed. Using virtual routers means that if one
fails because its server went down, another can be easily spun up to take its
place. Virtual routers are also essential in many cloud applications. Public
cloud providers are understandably unenthusiastic about your sending your techs
to install routers in their data centers, but you may need a virtual router
there if you want to use virtual networking and SD-WAN optimally. For this type
of cloud virtual routing, make sure your virtual router is compatible with the
virtual network or SD-WAN technology you’re using.
Overcoming the roadblocks to passwordless authentication
There are a variety of roadblocks associated with moving to passwordless
authentication. Foremost is that people hate change. End users push back when
you ask them to abandon the familiar password-based login page and go through
the rigamarole of registering a factor or device required for typical
passwordless flows. Further, the app owners will often resist changing them to
support passwordless flows. Overcoming these obstacles can be hard and
expensive. It can also be exacerbated by the need to support more than one
vendor’s passwordless solution. For example, most passwordless solutions pose
app-level integration challenges that require implementing SDKs to support even
simple flows. What happens if you want to support more than one solution? Or use
your passwordless solution as both a primary identity and authentication
provider and a step-up authentication provider? Or you want to layer in
behavioral analytics? There is a way to address these human and technical
challenges standing in the way of passwordless adoption using orchestration.
Although common in virtualized computing stacks, orchestration is a new concept
in identity architectures.
Obsolescence management for IT leaders
Obsolescence will always be a by-product of continuous technological advances.
The best way to improve cyber security and reduce downtime risks is to prepare
effectively and take proactive steps to manage obsolescence. With a proactive
obsolescence management plan in place, such as a cloud-first approach,
businesses can track the lifespan of products. This ensures that IT and
operational technology are always protected, improving productivity and
reducing costs. To plan for the future, mid-size businesses should carry out
an assessment of current infrastructure to understand the components of the IT
and operational technology landscape and how these systems interact. Vendors
will often publish end-of-life dates for hardware and software at least twelve
months in advance. IT managers should look at how much they already spend on
maintenance and whether downtime has occurred before. Understanding the risks
can also help businesses make more informed decisions about their equipment.
Businesses should consider how the failure of a hardware or software component
will impact operations, costs and reputation, and whether the equipment is
compatible with the rest of the system.
The pitfalls of poor data management – and how to avoid them
One of the challenges is how differences in patient profile can drastically
change the costs associated with the same procedure. For example, a healthy
patient with no comorbidities can likely receive a colonoscopy at an
outpatient center. However, a patient with a medical condition such as
hemophilia would need that same colonoscopy performed in the more costly
hospital setting because of the complications that could potentially arise.
This variability makes providing accurate estimates complicated. One way to
potentially address this issue is to provide best-case and worst-case
estimates. Getting to the point where these estimates can be made in real
time, so that a procedure can safely continue when a complication arises
without the concern of being fined or not properly reimbursed, is key. Also,
while the regulations are well-intended, the reality is it is probably
unnecessary to have the specified level of price transparency for every
encounter. We need to focus on the most problematic events – those medical
episodes that bankrupt people because they had no idea what their
out-of-pocket costs would be.
Icelandic datacentres may lead the way to green IT
One of the main application areas where Icelandic datacentres make a lot of
sense is in artificial intelligence (AI). With the advancement of AI
methodologies such as unsupervised machine learning, for many applications, AI
training and inference now needs to occur in the same location – they need to
be colocated to facilitate iteration between the two processes. Foundational
AI models run for weeks or months to do a re-education, so running a full
training data set is very energy intensive. Businesses that depend on AI
models do training continuously to get different versions of the models. For
example, they might train for a specific customer who has a data set they want
trained against. ... A second type of application where Icelandic datacentres
make sense is in financial services. Although trading applications require
very low latency and are usually placed close to exchanges in edge or metro
locations, they depend on the output of larger, more compute intensive
applications. These applications use thousands of computers 24 hours a day to
run Monte Carlo simulations and Markov Chain analysis to make predictions
about market trends.
Automotive hacking – the cyber risk auto insurers must consider
Cyber exposures are a relatively new frontier for auto insurance. Traditional
risk considerations have revolved around liability or theft, but those have
evolved amid the increasingly connected landscape for vehicles. “We must
evaluate the types of losses happening and what’s causing those losses. Are
they related to malfunctions in a vehicle? Are they related to hacking? It’s a
challenge for insurers even to determine the ultimate cause of a loss,” said
Perfetto. “If there was an accident, and it wasn’t the driver’s fault per se
but more of a vehicle malfunction, that may not be easily attributed. If there
was a hacking incident, that might not be easy to discover.” ... “We have seen
data that supports reduction in accident frequency related to certain
technology added to a vehicle. But we have also seen the cost of replacing
some more advanced technologies increase. Something as simple as a rear end or
a minor dent in your bumper that used to be an easy and relatively inexpensive
item to fix has become much more costly,” Perfetto noted.
Are debt financings the new venture round for fintechs startups?
You have to plan ahead for venture debt. Put it in place relatively soon after
an equity financing. That way there is no adverse selection for the lenders;
everyone (founders, VCs and lenders) around the table is happy at that time.
If you try to put something in place with less than six months of cash, you
will not be able to get debt. If you put it in place after an equity round,
you can draw it down way into the future — that’s called a forward
commitment/drawdown. That gives the startup a lot of optionality. It’s super
important to understand all the terms. Often, founders don’t realize there are
things like funding MACs, investor abandonment clauses, etc. These terms can
be used by the lender to block the startup from either drawing down the money
or creating a default after the money has been drawn. Either way, the company
is in trouble and can’t count on the capital. So you really need to know your
lender, have your VCs know your lender and pay attention to your terms. This
is why we created the Sample Venture Debt Term Sheet, to explain all the
terms.
The cybersecurity skills gap is ‘not just about addressing headcount’
From a security perspective, I’m hoping an increase in connected systems will
lead to less human-error-related cyberattacks. This will largely revolve
around increasing API accessibility and integration. Not only do better
integrations allow for employees to do better, more efficient work, it also
enables a more secure infrastructure throughout your entire organisation. For
example, when APIs are accessible throughout the application ecosystem, this
allows for systems to be configured through code, helping us introduce
streamlined changes to configuration rather than having to go into specific
applications. From a security perspective, this enables us to do advanced
things like segregation of duty and activity monitoring at scale. These
benefits are a large part of why we prioritise connectivity and API
accessibility at Templafy, both in our own tech stack and our platform. We
know it not only benefits our own team, but also our customers.
IT leadership: Why adaptability matters
The rise of technology has incentivized industries to adapt in recent years.
Still, that push is becoming a pull as realities like The Great Resignation and
remote work push organizations to change how they interact with and relate to
their customers and employees. The return on investment of developing
adaptability in organizations comes from talent attraction and retention,
increased innovation, improved employee engagement – and potentially,
organizational survival. In the past, leaders have been able to draw from models
such as William Bridges’ Transitions to understand adaptability. But while these
approaches may help us to understand how a person adapts and what behaviors
leaders should expect as people move through change, few have explored the why.
And without that knowledge, it can be challenging for leaders to create
supportive, psychologically healthy workplaces that support people as they
adapt. Because adapt they must. The key to unlocking the potential of emotional
intelligence is first to understand the construct and then identify the areas
for development. The same goes for AQ.
Developer Experience vs. User Experience
Retaining developers requires more than first impressions. Just as good UX needs
to be evaluated, refined, and tested over time, good DX is an investment in the
long term. You won’t know how well you’ve succeeded without using analytics to
evaluate your DX and test changes. Monitoring your API helps you identify users
who have not been able to successfully make API calls, find patterns of success
and failure for developers, and see how different users are engaging with your
product over time. While tracking UX metrics is relatively straightforward for
products focused on end-users, DX metrics differ in important ways. You need to
develop a good strategy for API analytics so that you track relevant business
value metrics while avoiding vanity metrics. ... You need to understand DX when
you build products for developers so that you can attract developer users,
inspire their confidence and creativity, and support their increasingly complex
integrations over time. Building good UX and DX can be challenging, but with the
right analytics stack, you can monitor your API and use metrics to craft the
perfect API developer experience.
Quote for the day:
"Taking charge of your own learning is a
part of taking charge of your life, which is the sine qua non in becoming an
integrated person." -- Warren G. Bennis
No comments:
Post a Comment