Date   

Re: Upcoming category changes

Dan Kohn <dan@...>
 

Yep, we don't have any plans to change serverless right now, but feel free to make suggestions.

--
Dan Kohn <dan@...>
Executive Director, Cloud Native Computing Foundation https://www.cncf.io
+1-415-233-1000 https://www.dankohn.com

On Thu, Jul 12, 2018 at 1:13 PM, Joshua Thorngren <josh@...> wrote:
Hi Dan - thanks for the recap - looks good. 

Apologies if my question was covered in the call - but I'm wondering what the current thought on updating the reference architecture/landscape with regards to serverless is?  Is the plan to still keep serverless as a separate landscape category? 

On Wed, Jul 11, 2018 at 11:46 PM Dan Kohn <dan@...> wrote:
We had a good call today and are planning to make the following category changes. Please speak up in the next week if you don't agree, and in particular if you have a different suggestion.

The Service Management category will be broken up as follows:

REMOTE PROCEDURE CALL
Apache Thrift
Avro
gRPC
Netflix Ribbon

LOAD BALANCER
Avi Networks
Backplane
F5
HAProxy
Heptio Contour
NGINX
Traefik

API GATEWAY
3Scale
Ambassador
Kong
MuleSoft
Reactive Interaction Gateway

SERVICE MESH
Conduit
Envoy
Hystrix
Istio
Linkerd
Netflix Zuul
Turbine Labs Houston
Vamp

Moved to new Security & Compliance section next to Key Management: Open Policy Agent (OPA)
Moved to Application Definition: Open Service Broker API
Moved to Databases: Stolon

Separately, these 3 categories will be combined: Host Management/Tooling, Infrastructure Automation, Private Cloud. The new category will be called "Automation and Configuration".

Secure Images will be renamed to Security & Compliance, and Oracle Policy Automation will be moved there.

We're planning to add a new Chaos Engineering to Observability & Analysis
--
Dan Kohn <dan@...>
Executive Director, Cloud Native Computing Foundation https://www.cncf.io

--

Josh Thorngren

VP, Marketing | Twistlock

josh@...

+1-971-645-7736



Re: Upcoming category changes

Joshua Thorngren
 

Hi Dan - thanks for the recap - looks good. 

Apologies if my question was covered in the call - but I'm wondering what the current thought on updating the reference architecture/landscape with regards to serverless is?  Is the plan to still keep serverless as a separate landscape category? 


On Wed, Jul 11, 2018 at 11:46 PM Dan Kohn <dan@...> wrote:
We had a good call today and are planning to make the following category changes. Please speak up in the next week if you don't agree, and in particular if you have a different suggestion.

The Service Management category will be broken up as follows:

REMOTE PROCEDURE CALL
Apache Thrift
Avro
gRPC
Netflix Ribbon

LOAD BALANCER
Avi Networks
Backplane
F5
HAProxy
Heptio Contour
NGINX
Traefik

API GATEWAY
3Scale
Ambassador
Kong
MuleSoft
Reactive Interaction Gateway

SERVICE MESH
Conduit
Envoy
Hystrix
Istio
Linkerd
Netflix Zuul
Turbine Labs Houston
Vamp

Moved to new Security & Compliance section next to Key Management: Open Policy Agent (OPA)
Moved to Application Definition: Open Service Broker API
Moved to Databases: Stolon

Separately, these 3 categories will be combined: Host Management/Tooling, Infrastructure Automation, Private Cloud. The new category will be called "Automation and Configuration".

Secure Images will be renamed to Security & Compliance, and Oracle Policy Automation will be moved there.

We're planning to add a new Chaos Engineering to Observability & Analysis
--
Dan Kohn <dan@...>
Executive Director, Cloud Native Computing Foundation https://www.cncf.io

--

Josh Thorngren

VP, Marketing | Twistlock

josh@...

+1-971-645-7736


Upcoming category changes

Dan Kohn <dan@...>
 

We had a good call today and are planning to make the following category changes. Please speak up in the next week if you don't agree, and in particular if you have a different suggestion.

The Service Management category will be broken up as follows:

REMOTE PROCEDURE CALL
Apache Thrift
Avro
gRPC
Netflix Ribbon

LOAD BALANCER
Avi Networks
Backplane
F5
HAProxy
Heptio Contour
NGINX
Traefik

API GATEWAY
3Scale
Ambassador
Kong
MuleSoft
Reactive Interaction Gateway

SERVICE MESH
Conduit
Envoy
Hystrix
Istio
Linkerd
Netflix Zuul
Turbine Labs Houston
Vamp

Moved to new Security & Compliance section next to Key Management: Open Policy Agent (OPA)
Moved to Application Definition: Open Service Broker API
Moved to Databases: Stolon

Separately, these 3 categories will be combined: Host Management/Tooling, Infrastructure Automation, Private Cloud. The new category will be called "Automation and Configuration".

Secure Images will be renamed to Security & Compliance, and Oracle Policy Automation will be moved there.

We're planning to add a new Chaos Engineering to Observability & Analysis
--
Dan Kohn <dan@...>
Executive Director, Cloud Native Computing Foundation https://www.cncf.io
+1-415-233-1000 https://www.dankohn.com


Re: How to reorganize Service Management

Chip Childers
 

On Fri, Jun 29, 2018 at 7:20 AM Dan Kohn <dan@...> wrote:
API GATEWAYS
3Scale
Ambassador
Kong
MuleSoft
Open Service Broker API

Open Service Broker API isn't an API gateway. I was going to suggest "service discovery", but it doesn't do runtime service discovery or even touch data path. Looking at the other categories it feels like "other" may be the best option. That, or "service catalog" becomes a category. 

--
Chip Childers
CTO, Cloud Foundry Foundation
1.267.250.0815


Re: How to reorganize Service Management

Lee Calcote
 

This one is a bit of a challenge. A few notes that may help clarify:

  • Turbine Labs is the vendor, Houston is their service mesh.
  • Yes, API Gateways are the closest in resemblance to Service Meshes, load-balancers next in line in terms of resemblance.
  • While OPA’s canonical use cases are security-oriented, as a policy engine it’s broadly capable of evaluating policies of many domains (non-security domains).
  • Hysterix should be booted from the service mesh category. It’s a client library, missing a control plane - one of the distinguishing characteristics of a service mesh.
    • While it might be painful to relabel, the same qualification applies to Envoy unto its own is falls into the Load Balancer category.

- Lee

On Jun 29, 2018, at 6:20 AM, Dan Kohn <dan@...> wrote:

https://landscape.cncf.io/grouping=landscape&landscape=service-management

The service management sub-category of the landscape is, IMHO, one of the most confusing.

It appears to be a grab bag of the following:

API GATEWAYS
3Scale
Ambassador
Kong
MuleSoft
Open Service Broker API

SERVICE MESH
Conduit
Envoy
Hystrix
Istio
Linkerd
Netflix Zuul
Turbine Labs
Vamp

LOAD BALANCER
Avi Networks
Backplane
F5
HAProxy
Heptio Contour
NGINX
Traefik

RPC
Apache Thrift
Avro
gRPC
Netflix Ribbon

OTHER
Open Policy Agent (OPA)
Reactive Interaction Gateway
Stolon

I would argue that Load Balancers are ultimately a subset of Service Mesh functionality, so those two categories should be combined. It's also debatable whether API gateways are a subset of Service Meshes.

RPC at least seems different to me, although those 4 might best be moved to the Streaming category: https://landscape.cncf.io/grouping=landscape&landscape=service-management,streaming

If we rename "Secure Images" to "Security and Policy", then Open Policy Agent (OPA) would probably belong there.

Your thoughts?
--
Dan Kohn <dan@...>
Executive Director, Cloud Native Computing Foundation https://www.cncf.io
+1-415-233-1000 https://www.dankohn.com


How to reorganize Service Management

Dan Kohn <dan@...>
 

https://landscape.cncf.io/grouping=landscape&landscape=service-management

The service management sub-category of the landscape is, IMHO, one of the most confusing.

It appears to be a grab bag of the following:

API GATEWAYS
3Scale
Ambassador
Kong
MuleSoft
Open Service Broker API

SERVICE MESH
Conduit
Envoy
Hystrix
Istio
Linkerd
Netflix Zuul
Turbine Labs
Vamp

LOAD BALANCER
Avi Networks
Backplane
F5
HAProxy
Heptio Contour
NGINX
Traefik

RPC
Apache Thrift
Avro
gRPC
Netflix Ribbon

OTHER
Open Policy Agent (OPA)
Reactive Interaction Gateway
Stolon

I would argue that Load Balancers are ultimately a subset of Service Mesh functionality, so those two categories should be combined. It's also debatable whether API gateways are a subset of Service Meshes.

RPC at least seems different to me, although those 4 might best be moved to the Streaming category: https://landscape.cncf.io/grouping=landscape&landscape=service-management,streaming

If we rename "Secure Images" to "Security and Policy", then Open Policy Agent (OPA) would probably belong there.

Your thoughts?
--
Dan Kohn <dan@...>
Executive Director, Cloud Native Computing Foundation https://www.cncf.io
+1-415-233-1000 https://www.dankohn.com


[cncf-toc] Updating the CNCF Reference Architecture

William Choe <william@...>
 

I missed the mtg, so here are a few comments/suggestions; 

1.  Leveraging the zero trust model, should there be an identification process for all entities

2.  On slide 6, policy is referenced.  How about an explicit call out for auth N, Z? 

3.  On slide 7,  it seems there should be a Cloud Native - Security segment.  

4.  Add security management for credentials, keys and certificates.  

Regards, 
william


On Wed, Jun 13, 2018 at 6:31 AM, Dan Kohn <dan@...> wrote:
In November 2016, the TOC approved this reference architecture: https://docs.google.com/presentation/u/1/d/1uMw2wkK0ubmc3khxqIuxK_rLK_wN89tNCnK7gDmTGR8/

It's best known as the basis for the groupings the Cloud Native Landscape: https://github.com/cncf/landscape#current-version

We've kicked off a mailing list and monthly phone call at 11 AM PT today to discuss revisions.

If you have views about how the architecture should change, please participate on the cncf-reference-architecture@... and the call. (And if you reply to the message, please remove cncf-toc@... from the response.) Any outputs from this effort would then go for a vote to the full TOC.

Thanks.

Topic: CNCF Reference Architecture
Time: Every month on the Second Wednesday, 11:00 AM Pacific Time (US and Canada)
Join from PC, Mac, Linux, iOS or Android: https://zoom.us/my/cncfrefarch

Or Telephone:
    Dial(for higher quality, dial a number based on your current location): 
        US: +1 669 900 6833  or +1 646 558 8656  or +1 877 369 0926 (Toll Free) or +1 855 880 1246 (Toll Free)
    Meeting ID: 304 550 1079
    International numbers available: https://zoom.us/u/cGgwaajOj
NOTE: Please use *6 to mute/un-mute your phone during the call.
--
Dan Kohn <dan@...>
Executive Director, Cloud Native Computing Foundation https://www.cncf.io
+1-415-233-1000 https://www.dankohn.com





Re: App Dev / App Ops

Craig Peters
 

Matt,

Thanks for staring the thread. Definitely that is where my interest lies.

Best,
Craig

On Wed, Jun 13, 2018 at 12:04 PM Matt Farina <matt@...> wrote:
On the call today we talked organizing around different parts of the ref architecture.

I wanted to share that I'm interested in the App Dev and App Ops portions. Those are the places where I can be most useful. Is anyone else interested in this space?

Cheers,
Matt Farina

--
Craig Peters | Director of Product, Partners | JFrog
http://www.jfrog.com | m: +1 925-639-0804 | skype: craig.l.peters


App Dev / App Ops

Matt Farina
 

On the call today we talked organizing around different parts of the ref architecture.

I wanted to share that I'm interested in the App Dev and App Ops portions. Those are the places where I can be most useful. Is anyone else interested in this space?

Cheers,
Matt Farina


User Roles

Matt Farina
 

There are a few, at least, roles that came to mind from the discussion today. The same person could do them but in some larger organizations they may be different people:

  • App Developer - builds the applications and is focused on the business logic of their application. An application could likely be run in more than one environment.
  • App Operator - takes an application and sets it up to operate in an environment. This could be a platform, such as Kubernetes, but also deals with varying environments such as test/QA/prod or varying global locations.
  • Infra Architect - figures out which pieces will be used in the platform. For example, chooses between Prometheus, Monaska, or something else for monitoring. Another example would be to choose storage technologies.
  • Infra Operator - deals with operating the infrastructure day in and day out
Each of these will have slightly different concerns and the decisions on tools to use in each case will be different.

This is inspired by the user profiles we generated for Helm.

Cheers,
Matt Farina


Suggested updates

Dan Kohn <dan@...>
 

I'll start off by suggesting that I think the architecture is largely fine and doesn't need a lot of changes. Rather than wholesale re-arrangement, I would prefer to look at some of the boxes with more items and see if there is a clear way to subdivide them.


"Orchestration & Management - Service Management (26)" is somewhat of a hodgepodge with service meshes (Envoy, Linkerd), messaging (gRPC), and service discovery (Open Service Broker API) all mixed together.

"App Definition and Development - Database and Data Warehouse (39)" is the largest subcategory with 39 but I don't see an obvious further division to do.

"Observability and Analysis - Monitoring (41)" is the largest but I also don't see an obvious breakdown.

I'd also like to consider whether "Cloud - Private (5)" and "Provisioning - Host Management/Tooling (9)" and "Provisioning - Infrastructure Automation (8)" should be combined or re-organized in some way.
--
Dan Kohn <dan@...>
Executive Director, Cloud Native Computing Foundation https://www.cncf.io
+1-415-233-1000 https://www.dankohn.com


Updating the CNCF Reference Architecture

Dan Kohn <dan@...>
 

In November 2016, the TOC approved this reference architecture: https://docs.google.com/presentation/u/1/d/1uMw2wkK0ubmc3khxqIuxK_rLK_wN89tNCnK7gDmTGR8/

It's best known as the basis for the groupings the Cloud Native Landscape: https://github.com/cncf/landscape#current-version

We've kicked off a mailing list and monthly phone call at 11 AM PT today to discuss revisions.

If you have views about how the architecture should change, please participate on the cncf-reference-architecture@... and the call. (And if you reply to the message, please remove cncf-toc@... from the response.) Any outputs from this effort would then go for a vote to the full TOC.

Thanks.

Topic: CNCF Reference Architecture
Time: Every month on the Second Wednesday, 11:00 AM Pacific Time (US and Canada)
Join from PC, Mac, Linux, iOS or Android: https://zoom.us/my/cncfrefarch

Or Telephone:
    Dial(for higher quality, dial a number based on your current location): 
        US: +1 669 900 6833  or +1 646 558 8656  or +1 877 369 0926 (Toll Free) or +1 855 880 1246 (Toll Free)
    Meeting ID: 304 550 1079
    International numbers available: https://zoom.us/u/cGgwaajOj
NOTE: Please use *6 to mute/un-mute your phone during the call.
--
Dan Kohn <dan@...>
Executive Director, Cloud Native Computing Foundation https://www.cncf.io
+1-415-233-1000 https://www.dankohn.com


CNCF Reference Architecture Update

Dan Kohn <dan@...>
 

Last chance to register with the Doodle poll. Please do so.
--
Dan Kohn <dan@...>
Executive Director, Cloud Native Computing Foundation https://www.cncf.io
+1-415-233-1000 https://www.dankohn.com

---------- Forwarded message ----------
From: Dan Kohn <dan@...>
Date: Mon, May 21, 2018 at 7:33 PM
Subject: CNCF Reference Architecture Update
To: cncf-toc@...


In November 2016, the TOC approved this reference architecture.

It's best known for being the basis of the cloud native landscape and interactive landscape.

Ken Owens is starting an effort to revise the document. It is unlikely to need to be an actual working group, but we created a (public) mailing list to have a place to discuss things. We also have a Doodle poll for when to hold a twice a month meeting.

If you're interested in participating, please join the mailing list.
--
Dan Kohn <dan@...>
Executive Director, Cloud Native Computing Foundation https://www.cncf.io
+1-415-233-1000 https://www.dankohn.com

21 - 33 of 33