Hi, I mentioned at next meeting presenting the work and ideas which we have been trying out for s390x around the image pull challenges and future ideas. With holidays in China for our main presenter and KubeCon next week I hope it will be OK for us to present this and some ideas for possible extensions to shim api etc the following week? So Thursday 13th May. Thanks James Magowan IBM Hyper Protect Containers (ZaaS Kubernetes) Mobile: 44-7808-247946 | Phone: 44-1962-817235 E-mail: MAGOWAN@uk.ibm.com Hursley Park Hursley, Hampshire SO212JN United Kingdom IBM United Kingdom Limited Registered in England and Wales with number 741598 Registered office: PO Box 41, North Harbour, Portsmouth, Hants. PO6 3AU From: Samuel Ortiz <sameo@linux.intel.com> To: Ariel Adam <aadam@redhat.com>, kata-dev <kata-dev@lists.katacontainers.io> Cc: Peng Tao <bergwolf@antfin.com> Date: 29/04/2021 15:41 Subject: [EXTERNAL] Re: [kata-dev] Community use cases meeting Thanks everyone for joining today's meeting, very interesting discussions. Please note that we decided to move to a weekly cadence, we have a lot to talk about it seems. Below you will find the meeting minutes, extracted from the rolling meeting notes document [1] * Current set of pull requests related to enablement (3 PRs) - How do we want to proceed with those PRs? - Samuel thinks that merging the TDX could help overall (a few more comments to request). The idea would be to rebase the 2 others on top of this one. Any objections? - We agree to merge the framework first and pile the other patches on top of it :-) * Agent API - discussion on what API restrictions we need (build time option or not) - What is the risk of having this code in the image however just not used? - It could be a runtime configuration for the agent however still part of the image (inside the initrd) - We may have benefits from restricting the APIs (reducing attack surface) - Could we limit things to always getting all the pod configuration at once (and nothing else?) - So the idea would be to drop those APIs for all use cases * Container image service offload - PEV raksh codebase -> https://github.com/IBM/raksh-agent/commit/14b52fae8c7bc3ce4b226703efdb37749c... * Testing for confidential computing - and general expectations - A big question regarding how to test the different platform so it’s something accessible upstream (SEV and TDX) - Intel will provide HW for testing as part of the kata CI - Can AMD also provide HW for the kata CI? - Sandeep pointed out they have the HW for this however the question is how to make the HW available for kata CI testing - What tests would we want to run in this case? - Samuel will pull Salvador to look into this - We may start by looking at the integration tests and later on E2E integration with k8s * Action Items - Ariel to change the name of the meeting to confidential computing - Ariel to change the meeting cadence to a weekly one - Samuel to send a doodle to poll for meeting time changes [1] https://docs.google.com/document/d/1E3GLCzNgrcigUlgWAZYlgqNTdVwiMwCRTJ0QnJhL... On Wed, Apr 28, 2021 at 11:23:23PM +0200, Samuel Ortiz wrote:
All,
A friendly reminder that we will have our weekly use case meeting tomorrow (04/29) at 12:00 UTC. See you all tomorrow!
Cheers, Samuel.
On Mon, Apr 19, 2021 at 08:12:34PM +0300, Ariel Adam wrote:
On Mon, Apr 19, 2021 at 2:25 PM James Magowan <MAGOWAN@uk.ibm.com> wrote:
Hi Ariel/Samuel,
Moving Confidential computing Use Case meeting is fine for me.
Given recent fun with summer savings time and the dates/times in the thread.
Can we clarify that meetings will be? every 2 weeks at 12:00 UTC
next meetings Thursday 29th April, Thursday 13th May etc.
I've moved the meeting 1 week forward to April the 29th and verified that the only invitee is kata-dev mailing list.
The meeting will be at 12:00 UTC and happen every 2 weeks. I will also make sure to send a reminder a day before the meeting (every 2 weeks).
Lastly, I'm reminding everyone that this time slot is mainly for the confidential computing use case and the CI/CD and performance use cases will be discussed at other times.
Thanks.
Thanks,
James *Mobile:* 44-7808-247946 | *Phone:* 44-1962-817235
*E-mail:* *MAGOWAN@uk.ibm.com* <MAGOWAN@uk.ibm.com> Hursley Park
Hursley, Hampshire SO212JN
United Kingdom
IBM United Kingdom Limited
Registered in England and Wales with number 741598
Registered office: PO Box 41, North Harbour, Portsmouth, Hants. PO6
3AU
From: Ariel Adam <aadam@redhat.com> To: Samuel Ortiz <sameo@linux.intel.com>, Eric Ernst < eric.g.ernst@gmail.com>, Peng Tao <bergwolf@antfin.com>, kata-dev < kata-dev@lists.katacontainers.io> Date: 18/04/2021 18:42 Subject: [EXTERNAL] Re: [kata-dev] Community use cases
------------------------------
Since you're leading the confidential computing use case which we agreed Thursday's time slot would be used for then to me that makes sense.
Does anyone object to moving the meeting by 1 week (as usual every 2 weeks from there)?
Also, a few people pointed out that they don't see the meeting or still see it every 2 weeks instead of 1 week. We basically sent an invite to the kata-dev for every 2 weeks so if anyone has a better idea on how to do that please let us know :-).
Thanks.
On Sun, Apr 18, 2021 at 8:32 PM Samuel Ortiz <*sameo@linux.intel.com* <sameo@linux.intel.com>> wrote:
Ariel,
Next week's meeting (04/21), I will be out of the office and won't be able to attend the meeting. Would anyone object if we shift that one to 04/28 instead ?
Cheers, Samuel.
On Wed, Apr 14, 2021 at 01:27:54PM +0300, Ariel Adam wrote:
Hi everyone. Going back to this mail thread. Talking with Samuel it will be hard for him to move the time slot for the confidential computing use case (currently on Thursday every 2 weeks 8AM EST).
What we suggest is the following:
- *Confidential computing* - will keep the current timeslot,
ping
Samuel directly if you would like to attend and this is a
you
- *CI/CD use case* - in practice it's only RH and IBM cloud
work mainly through mails (and ping me if you are also interested in
will this)
- *Performance isolation* - Tao will work on a different time slot to fit people who would like to attend as well (such as Eric)
As pointed, we will ask to provide a status on the use cases
during the AC meetings if time permits.
Comments are welcomed.
Thanks.
On Wed, Apr 7, 2021 at 5:08 PM Eric Ernst <*eric.g.ernst@gmail.com* <eric.g.ernst@gmail.com>> wrote:
Makes sense to me, and I appreciate the change!
Eric
On Apr 7, 2021, at 6:26 AM, Ariel Adam <*aadam@redhat.com* <aadam@redhat.com>> wrote:
Given that Tao agrees this makes sense I checked with Samuel as well.
What we suggest is that each use case leads (Ariel, Samuel, Tao) will send a separate doodle poll for their use case. That way each of the use case teams can decide on a time that fits the majority of participants and also on the cadence of the meetings (every 1 or 2 weeks).
We will also propose to give a short update on the progress of
different use cases on the weekly AC meeting (if time permits).
Let us know if anyone has any objections before we send out the doodle polls.
Thanks.
On Wed, Apr 7, 2021 at 3:03 PM Peng Tao via kata-dev < *kata-dev@lists.katacontainers.io* <kata-dev@lists.katacontainers.io>> wrote:
Thanks for bringing it to the list Fabiano! I am interested in
performance isolation use case and would love to see US folks
meeting please problem for thus we progress the the present
their thoughts and ideas too. So I am all for it to split the use case meeting and relocate to a time that works better for US folks.
Cheers, Tao
On 2021/4/7 16:26, Fabiano Fidêncio wrote: > Yesterday, Apr 6th, during the Architecture Committee meeting, we > discussed the possibility of having the community use cases meeting > split, and I assigned myself the action item to create a doodle and > send an email. > > Well, let me start saying that I may have failed in the doodle thing, > but I still am dropping the email with what I heard as a feedback and > what I would like to suggest. > > I got the feedback, from more than one person, that the time for the > community use cases meeting is not exactly optimal for everyone, and > that is right. For instance, it may be too early for folks in some > parts of the US, while it is still too late for folks in Australia. > > With this in mind, I would like to ask whether there is the interest > to split the meeting, allowing then the involved parts to talk to each > other in a time that suits them best. > > So, this message is for everyone, but here I am explicitly looking for > feedback from folks interested in: > * K8S CI/CD pipeline with kata containers > - Here we have people interested from the US and Australia, so maybe > a late meeting in the US time zone would benefit the folks in > Australia. > * Performance isolation > - Here we have people interested from the US and China, so maybe a > late meeting in the US time zone would benefit the folks in China. > > If the parts involved in this have the interest to do so, let me know > and then I will try to organise a doodle, set an official time, etc, > etc, etc. > > We, as the community facilitating this communication, are here to > facilitate the communication in a way that it will be beneficial for > all the involved parts. > > Best Regards, >
-- Into something rich and strange.
_______________________________________________ kata-dev mailing list *kata-dev@lists.katacontainers.io* <kata-dev@lists.katacontainers.io> * http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev*
< http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev
_______________________________________________ kata-dev mailing list *kata-dev@lists.katacontainers.io* <kata-dev@lists.katacontainers.io> * http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev*
< http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev
_______________________________________________ kata-dev mailing list *kata-dev@lists.katacontainers.io* <kata-dev@lists.katacontainers.io> * http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev*
< http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev
_______________________________________________ kata-dev mailing list kata-dev@lists.katacontainers.io
http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev
Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with
number
741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
_______________________________________________ kata-dev mailing list kata-dev@lists.katacontainers.io
http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev _______________________________________________ kata-dev mailing list kata-dev@lists.katacontainers.io http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev Unless stated otherwise above: IBM United Kingdom Limited - Registered in England and Wales with number 741598. Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU