Zero trust paper

[mrsabath]

Publishing CNCF Zero Trust White Paper.
This PR replaces the original #1229 as requested by TAG Security reviewers.

The Markdown is based on Google Doc: https://docs.google.com/document/d/18w243JXhGxNgJk0SS1gT-LWUw29F8KFrc1D1F9dvfhw/edit?usp=sharing

CNCF issue: #950

This is still a draft. I just started the conversion and it requires more Markdown work and cleanup

[netlify]

✅ Deploy Preview for tag-security ready!

Name	Link
🔨 Latest commit	b7a72ee
🔍 Latest deploy log	https://app.netlify.com/sites/tag-security/deploys/6735725e1b95c10008e74776
😎 Deploy Preview	https://deploy-preview-1395--tag-security.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site configuration.

[TheFoxAtWork]

I dont believe this paper is ready for TOC liaison approval at this time. There are several formatting issues with the paper (such as inconsistent new lines) and the content of the paper could benefit from a significant amount of brevity in many areas that repeat or rephrase prior content. There are several areas of content that make leaps in logic or dont quite convey the meaning of a particular topic. Additionally, several sections don't provide the reader with new or meaningful guidance and examples in order to apply Zero Trust to their environments using cloud native tooling.

I have not read the paper in its entirety due to the number of areas that would greatly benefit from additional refinement by the authors. The paper is sorely needed, and i appreciate the time and effort by the contributors in putting this draft together for an initial review and feedback.

[TheFoxAtWork]

I would argue that none of the concepts in Zero Trust are well-worn and that it is still a challenge to apply them to any system, thus why there remains gaps in approach Zero Trust with cloud native in mind. If it were so well-worn, this paper would have been published in 2022 and many government agencies would not be releasing their ZT strategies in 2024.

Suggested change

	While many of the well-worn concepts behind Zero Trust apply to *any* system, there remains a gap with regards to discussing Zero Trust from a Cloud Native perspective. This document seeks to codify the philosophy alongside an ideal design for implementing it in a Cloud Native system.
	Zero Trust concepts have been around for a while, however there continues to be a gap in how industry can effectively adopt and apply them across all systems and applications in use, Cloud Native is no exception. This document seeks to codify the philosophy alongside a recommended design for implementing Zero Trust principles in Cloud Native systems and applications.

[davidhadas]

I would argue that the concepts are well-worn, while I agree that it is still a challenge to apply them to any system :)
I.e. Many of the concepts defining how to build a ZTA are here with us for quite a while - not suggesting that we have the tools and mechanisms to implement them.

At the same time, I am also perfectly ok with the suggested change - so I would vote to accept it as is.

[TheFoxAtWork]

Fair point.

[mrsabath]

@TheFoxAtWork are you OK with us keeping it as is, or you like this to be changed per your suggestion?
BTW I tried to find you at the conference to say "hello", but no luck...

[TheFoxAtWork]

perhaps it is later, but this doesnt detail why active observers are being recommended, it just introduces the concept. Why are these beneficial specifically in cloud native for achieving Zero Trust and not in the broader technology ecosystem? Do they fill a un-noted gap that is critical to ZT implementation? Highly recommend adding in the illuminating moment that drives us towards active observers here so the reader is prepared for its more robust introduction later.

[davidhadas]

@TheFoxAtWork, We are not stating or suggesting that active observers are beneficial specifically in cloud native for achieving Zero Trust - these are beneficial in the broader technology ecosystem.

We do state that “Active Observers,” are a previously unnamed category of tools - i.e. we help name and better define such tools and help explain how they are being used. There are mentioning of such capabilities in previous ZT docs, we take this one step further.

[TheFoxAtWork]

The heading/titling structure here does not match with that of Always Verify, which has a secondary phrase enhancement. Recommend either eliminating the enhancing phrase or adding an enhancing phrase to the Assume a Breach. It doesnt appear to add substantial value however and given the table is quite simple in its content, you can increase its value by providing a more clear example of what is meant or pose questions to the reader to have them consider the context:

Assume a breach: Clients will send malicious requests. In the case a client begins sending malicious requests, are you servers capable of authenticating the client and verifying its authorization to connect prior? In the event that a valid, authorized client is compromised, are your servers capable of validating the input in the event it is malformed? what about sending too many requests? Do you have detection rules in place to identify unusual behaviour by clients?

[TheFoxAtWork]

I would further recommend linking each item in the table to the corresponding H3 below in the event a reader wishes to jump down into that content.

[davidhadas]

It seems repetitive to spell out each one in the table and also in the below H3 sections.
I believe the second column help connect the principles to the NIST Tenet which is part of covering where these principles come from.

@mrsabath can we manually add links to the table?
And do we also need it in the doc?

[mrsabath]

Yes, we can do them manually, but we have to make sure that any changes in this document are also reflected in the Google Doc that would be used for rendering the PDF, so I suggest ONLY ONE PERSON does the updates in both places, otherwise we will get out of sync quickly. I am happy to do them, but I need to know exactly what to update. Since I am on KubeCon too, the edits might be delayed...

[davidhadas]

@mrsabath
We continue to rely on you for making the changes and keeping the two documents inline.

Where a reviewer made a specific edit suggestion, I am trying to help by reading the text with the suggestion and adding a thumb up for support.

Where a reviewer commented without a specific edit suggestion, I am trying to either add my own specific edit suggestion or commenting that it may be preferable not do a change.

[mrsabath]

I realized what you did @davidhadas and thank you for that. I really appreciate your help with managing the edits. I think I merged all the suggestions that were approved by you unless they require some further explanation or investigation.

[mrsabath]

Like this one: #1395 (comment)
Are we keeping it as is?

[davidhadas]

I think the text on that comment can stay as is, but lets see if others will also see a need for change and what that change may be.

[TheFoxAtWork]

I'm reading through these and trying to understand why we've chosen to talk about these challenges at a high level. We're not detailing anything about zero trust to the benefit of the reader.

for example, in the event a service is exploited, what specific capabilities, features, and tooling do we as cloud native security professionals recommend? Runtime detection tools assist in identifying if a host or workload was exploited. Tools that provide remote attestation have the ability to measure the integrity of files to detect if it has been tampered with.

these are specific recommendations we should be providing throughout this area of the paper that provide value to the reader.

[davidhadas]

We cover this later in the paper -

Here we spell out the principles to follow (adding one more detail to the overall ZT mindset the reader should maintain) - we suggest that the reader would consider that each and every service running on the CN may be exploited.

Later in the doc we discuss how should the reader detect such a situation and what should happen next.

[davidhadas]

@TheFoxAtWork - Will you be ok with keeping this text as is given my comment above?