Re: K8S as a cognitive architecture for clinical decision support?

[Michael Taufen]

Probably not the right forum, as this list is geared towards development of the system itself, rather than more abstract discussion. Do we have something like a kubernetes-misc list?

You raise an interesting idea. Some of the higher-level ways of thinking about problems may be useful across fields, and so forth.

My 2c on mapping the models together, which can maybe give you a better idea of how Kubernetes works:

Your diagnosis-goal unit sounds like state that would be represented by some object represented in the Kubernetes master's database (etcd).

In general, state (like the variables you mention) that determines how containers are orchestrated is stored in "API objects" in the master's database.

A controller would be the component that observes this state (and potentially other state, for example) and responds to it by taking action.

One of the actions the controller might take is to launch a container (a process), which performs some additional operations to bring about change.

The controller may also stop containers that are no longer necessary, again based on state that it observes.

Services are an abstraction that give containerized applications stable addresses, so that users can continue to use the same point-of-access

even if some of the containers that serve the application change (deleted, added, etc.). Think of a Service like a storefront, the same

store is still there even when employees are hired/fired.

Hope that helps,

Mike

On Mon, Aug 21, 2017 at 12:07 PM, Steven Merahn <sme...@medcanto.com> wrote:

Apologies to all if this post is inappropriate for this forum...I know nothing about coding or distributed systems or any of the stuff you all do.....

However, as a physician who has worked in medical problem solving, I stumbled upon Kubernetes and became slightly obsessed with the idea of applying the container/orchestration concept to medicine as an information architecture.  

I may have completely misinterpreted these concepts, but if we call a Service a diagnosis-goal unit (Type II diabetes = improve glucose processing) then all of the approaches to achieve goal (lifestyle nutrition, medication) would be a combination of an action (application) and the evidence for it (the library):  a container.  And there may be a number of containers that are relevant to a particular Service....these containers need to be 'orchestrated' based on a a number of factors, including patient specific and environmental variables.  There are additional parameters for when a particular container would be added or applied. Containers have lifecycles...based on the emerging evidence and medical literature; some containers may need to go away and new containers developed, along with the new 'score' for orchestrating their interactions (serial, parallel use cases, etc).

Applying these concepts as the basic information architecture of medicine, would allow for a level of concept integration that does not currently exist in medicine and the body of knowledge could be maintained in an an entirely new way....and be an amazing foundation to reimagine medical education...

I have spent decades exploring new ways to help doctors think and solve problems -- and believe that we need to explore and transfer concept from outside of medicine if we're really going to support productive evolution of healthcare.

Comments?  Be gentle please.

Steven

Steven Merahn, MD

--
You received this message because you are subscribed to the Google Groups "Kubernetes developer/contributor discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to kubernetes-dev+unsubscribe@googlegroups.com.
To post to this group, send email to kubernetes-dev@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-dev/f21fe5b2-bb41-45e1-85a2-26bd1f4dfadc%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Michael Taufen

Google SWE

[Steven Merahn MD]

thanks Mike — the controller concept fills an important gaps in my knowledge…

I’m going to explore a bit more!

S

Steven Merahn MD
sme...@medcanto.com
917-689-8954
linkedin.com/in/stevenmerahn
@modernMD

[Sarah Novotny]

perhaps kubernetes-users?

To post to this group, send email to kuberne...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-dev/f21fe5b2-bb41-45e1-85a2-26bd1f4dfadc%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Michael Taufen

Google SWE

--
You received this message because you are subscribed to the Google Groups "Kubernetes developer/contributor discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to kubernetes-dev+unsubscribe@googlegroups.com.
To post to this group, send email to kubernetes-dev@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-dev/CAHJadQ%2BhT3iVTLifZqvMfO%2BeMUC5Dwe3dnOx8r-%3DFLNeu0zb1A%40mail.gmail.com.