Long Term Support Working Group proposal

[Timothy Pepper]

The recent 1.12 release renewed community commentary that the project drops support of old releases too quickly.  Any support discussion dovetails into the “What is Kubernetes? Kernel? Distribution?” discussion.  With cloud provider and other monolith splitting you get a wider Kubernetes software bill of materials, which makes more complex test, validation, documentation, upgrades.  There’s even a question of whether a shorter release cycle might be useful too or instead of long term support.  So many questions, but we don’t arrive at clear answers or actions without a more formal focus. 

 

So we propose a Long Term Support Working Group (WG LTS) to provide a cross SIG location for focused collection of stakeholder feedback.   [WG LTS is simply shorter than “WG To LTS Or Not To LTS” or “WG What Are We Releasing And Why And How Is It Best Integrated, Validate, And Supported”, but should NOT be read in that shortness to imply LTS is the foregone conclusion.]

 

What would this working group aim to achieve?

• We hope for involvement from external users/operators, vendors, distributors, and hosting providers and also from our project internal SIGs, especially Cluster Lifecycle, Cloud Provider, Release, Architecture, but certainly this touches on all the SIGs. 
• With feedback, we can attempt an informed assessment on the tradeoffs of requirements and costs.  Is LTS the answer or not?  
  • If it is, we imagine the WG would disband having drafted a KEP, gotten it approved by the project broadly, and returned it to SIG Release to operationalize. 
  • If it is not, we perhaps find other answers to propose or at least have a better assessment documenting the state of the project’s support relative to its users’ requirements. 

Timeline:

 

We expect a best-case time line of discussions this fall, especially leveraging the upcoming two KubeCons, leading then to drafting of proposals and decisions in 2019.

 

Please see here for more information: https://docs.google.com/presentation/d/1-Z-mUNIs3mUi7AdP1KwoAVNviwKrCoo3lxMb5wzCWbk/edit?usp=sharing

Especially the latter section of “Background Details”.

 

Next Steps:

 

We expect a large number of vigorous objectors and supporters, and encourage folks to openly express their thoughts. We will catalog responses. And we all SHOULD keep an open mind here. The point of this proposed working group is specifically to capture feedback and assess whether there is a forward path for incremental improvement to the project’s support processes.

 

We’re looking here and now for lazy consensus that a working group is the correct Kubernetes project forum to explore this problem space further…which is hopefully in accordance with the current process for working group formation (just changed today).

If that is achieved we’ll look at setting up discussion forums/meetings/etc.

 

 

-- 

Tim Pepper, Dhawal Bhanushali, & SIG-Release

[Stephen Augustus]

+1 for exploration!

--
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-de...@googlegroups.com.
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/6CF00E6B-98AB-4C5C-A5E0-7F7E204DAC5E%40vmware.com.
For more options, visit https://groups.google.com/d/optout.

[Christoph Blecker]

I think this is a good topic to explore, and the right forum to explore it in (short lived, a cross cutting problem, etc).

+1 from me.

--

[Davanum Srinivas]

+1 from me. There will be lot of improvements that we can do to what we do today which will be good even if we do not end up doing the "full" LTS.

-- Dims

To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-dev/CADx2oGGSPq3DRTua4Dqz3OcRCL-ZSod5iXxmOrxby9Tgh6j_vA%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

--

Davanum Srinivas :: https://twitter.com/dims

[Sanjeev Kumar]

+1 to explore..

Regards, Sanjeev

To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-dev/CANw6fcEcS2iDZytxqeEqEO_S4oX-qXX0-LzYxekBMkx2sSu_6g%40mail.gmail.com.

[Tim St. Clair]

+1 from me, seeing how I was the instigator.

> --
> 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-de...@googlegroups.com.
> 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/6CF00E6B-98AB-4C5C-A5E0-7F7E204DAC5E%40vmware.com.
> For more options, visit https://groups.google.com/d/optout.



--

Cheers,
Timothy St. Clair

“Do all the good you can. By all the means you can. In all the ways
you can. In all the places you can. At all the times you can. To all
the people you can. As long as ever you can.”

[Aaron Crickenberger]

+1 for WG What Are We Releasing And Why And How Is It Best Integrated, Validated, And Supported

- aaron

[Stephen Augustus]

It does have a nice ring to it. :)

To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-dev/88dc0bbb-bc73-455c-ae7b-95e412b2dc67%40googlegroups.com.

[roger....@suse.com]

+1 for SUSE supporting this examination.

[Klaus Ma]

+100

— Klaus

 

---

收件人: kuberne...@googlegroups.com 代表 Timothy Pepper <tpe...@vmware.com>
发送时间: 星期二, 十月 16, 2018 03:09
收件人: kuberne...@googlegroups.com
抄送: Dhawal Yogesh Bhanushali
主题: Long Term Support Working Group proposal

 

[Michael Hausenblas]

+1 absolutely, goes without saying!

Cheers,
Michael

--
Michael Hausenblas
Ireland, Europe
http://mhausenblas.info/

-----Original Message-----
From: Timothy Pepper <tpe...@vmware.com>
Reply: Timothy Pepper <tpe...@vmware.com>
Date: 15 October 2018 at 20:09:04
To: kuberne...@googlegroups.com <kuberne...@googlegroups.com>
Cc: Dhawal Yogesh Bhanushali <dbhan...@vmware.com>
Subject:  Long Term Support Working Group proposal

> The recent 1.12 release renewed community commentary that the project drops support
> of old releases too quickly. Any support discussion dovetails into the “What is Kubernetes?
> Kernel? Distribution?” discussion. With cloud provider and other monolith splitting
> you get a wider Kubernetes software bill of materials, which makes more complex test,
> validation, documentation, upgrades. There’s even a question of whether a shorter
> release cycle might be useful too or instead of long term support. So many questions,
> but we don’t arrive at clear answers or actions without a more formal focus.
>
>
>
> So we propose a Long Term Support Working Group (WG LTS) to provide a cross SIG location
> for focused collection of stakeholder feedback. [WG LTS is simply shorter than “WG To
> LTS Or Not To LTS” or “WG What Are We Releasing And Why And How Is It Best Integrated, Validate,
> And Supported”, but should NOT be read in that shortness to imply LTS is the foregone conclusion.]
>
>
>
> What would this working group aim to achieve?
>

> * We hope for involvement from external users/operators, vendors, distributors, and

> hosting providers and also from our project internal SIGs, especially Cluster Lifecycle,
> Cloud Provider, Release, Architecture, but certainly this touches on all the SIGs.

> * With feedback, we can attempt an informed assessment on the tradeoffs of requirements

> and costs. Is LTS the answer or not?

> * If it is, we imagine the WG would disband having drafted a KEP, gotten it approved by the

> project broadly, and returned it to SIG Release to operationalize.

> * If it is not, we perhaps find other answers to propose or at least have a better assessment

> documenting the state of the project’s support relative to its users’ requirements.
>
> Timeline:
>
>
>
> We expect a best-case time line of discussions this fall, especially leveraging the
> upcoming two KubeCons, leading then to drafting of proposals and decisions in 2019.
>
>
>
> Please see here for more information: https://docs.google.com/presentation/d/1-Z-mUNIs3mUi7AdP1KwoAVNviwKrCoo3lxMb5wzCWbk/edit?usp=sharing
>
> Especially the latter section of “Background Details”.
>
>
>
> Next Steps:
>
>
> We expect a large number of vigorous objectors and supporters, and encourage folks to
> openly express their thoughts. We will catalog responses. And we all SHOULD keep an open
> mind here. The point of this proposed working group is specifically to capture feedback
> and assess whether there is a forward path for incremental improvement to the project’s
> support processes.
>
>
>
> We’re looking here and now for lazy consensus that a working group is the correct Kubernetes
> project forum to explore this problem space further…which is hopefully in accordance
> with the current process for working group formation
> (just changed today).
>
> If that is achieved we’ll look at setting up discussion forums/meetings/etc.
>
>
> --
> Tim Pepper, Dhawal Bhanushali, & SIG-Release
>

[Jintao Zhang]

+1 from me.

在 2018年10月16日星期二 UTC+8上午3:09:04，Tim Pepper写道：

[Pengfei Ni]

+1

在 2018年10月16日星期二 UTC+8上午3:09:04，Tim Pepper写道：

The recent 1.12 release renewed community commentary that the project drops support of old releases too quickly.  Any support discussion dovetails into the “What is Kubernetes? Kernel? Distribution?” discussion.  With cloud provider and other monolith splitting you get a wider Kubernetes software bill of materials, which makes more complex test, validation, documentation, upgrades.  There’s even a question of whether a shorter release cycle might be useful too or instead of long term support.  So many questions, but we don’t arrive at clear answers or actions without a more formal focus. 

[Daniel Smith]

I'd like to join. I've long claimed that the existing release cadence is the root cause of many of the longstanding problems of the Kubernetes project.

--

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-de...@googlegroups.com.
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/a11e1ad6-cb06-4311-9cbe-1f9bfd73f3c9%40googlegroups.com.

[marky.r...@gmail.com]

+1 from me. I would very much like to join in support of this

[Quinton Hoole]

+1 Great idea. I agree with lavalamp.

--

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-de...@googlegroups.com.
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/6f49d30d-651c-407e-b7ab-49e3dada7eaa%40googlegroups.com.

[Jordan Liggitt]

+1 to “WG What Are We Releasing And Why And How Is It Best Integrated, Validated, And Supported”.

This is also closely related to discussions sig-arch has been starting to have around stability, quality, consistency, and compatibility of our existing releases.

Jordan

To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-dev/CAL5aO5OY0Z7N5Oxs7m2nKz-Gb2M17F1g3Y0CTg%3DbTF_kHycaqg%40mail.gmail.com.

[Nick Young]

+1 from me too, fwiw.

To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-dev/CAMBP-pLwagP0-OKPnV4rJvTcR8o%2BRWnEmG2R%2BWXJJUfKoOEB3A%40mail.gmail.com.

[Clayton Coleman]

+1 - we owe it to ourselves to have this discussion.

To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-dev/CAMBP-pLwagP0-OKPnV4rJvTcR8o%2BRWnEmG2R%2BWXJJUfKoOEB3A%40mail.gmail.com.

[Timothy Pepper]

FYI all:

 

https://github.com/kubernetes/community/pull/2911

 

WG creation PR

 

-- 

Tim Pepper

Orchestration & Containers Lead

VMware Open Source Technology Center

To view this discussion on the web visit https://groups.google.com/d/msgid/kubernetes-dev/CAH16ShL2CpgrmtbjB%2BdJMoko%3Dj34nFavjg8CfWpL8YdvWvRr4Q%40mail.gmail.com.

[Timothy Pepper]

FYI we are in fact working on formally setting up a WG.

 

Slack:                                     #wg-lts

Group:                                  kubernet...@googlegroups.com

Meetings:                            biweekly Tuesday 9am Pacific (alternating weeks with SIG Cluster Lifecycle; will be adding one additional time for better global participation)

YouTube Playlist:              https://www.youtube.com/playlist?list=PL69nYSiGNLP13_zDqYfUjfLZ2Lu9a3pv-

 

k/community PR #2911 against sigs list

https://github.com/kubernetes/community/pull/2911

includes a draft charter.

 

Feedback and participation is VERY WELCOME!

 

-- 

Tim Pepper

Orchestration & Containers Lead

VMware Open Source Technology Center

 

From: <kuberne...@googlegroups.com> on behalf of Timothy Pepper <tpe...@vmware.com>

Date: Monday, October 15, 2018 at 12:09 PM
To: "kuberne...@googlegroups.com" <kuberne...@googlegroups.com>
Cc: Dhawal Yogesh Bhanushali <dbhan...@vmware.com>

--

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-de...@googlegroups.com.
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/6CF00E6B-98AB-4C5C-A5E0-7F7E204DAC5E%40vmware.com.