Confidential computing meeting minutes (2021/05/27)
All,
During our May 27th meeting, we continued dicussing the image service offload proposal, based on a more detailed description of the overall proposed architecture [1].
We had a large crowd attending the meeting, including key CRIO and containerd maintainers. Eventually, there was agreement from both of these projects that moving the image pulling layer in the guest makes sense for the confidential computing use case we're trying to build here. I'm expecting the discussions on the implementation details and steps to continue happening through the relevant issue [2], but we now know that the potential impact on CRIO and containerd is acceptable.
Tomorrow, I'm proposing that we discuss about the Host/Guest API and how/if we'd want to split it. The agenda is at the usual place [3].
Last but not least: Starting from tomorrow, this meeting will alternate between 2 time slots: 12 noon (UTC) and 3pm (UTC). Tomorrow's (June 3rd) meeting starts at noon UTC.
Cheers, Samuel.
[1] https://docs.google.com/presentation/d/1469nSRFtlHMSDDDWVLj0i21dR9M_3SO76ehQ... [2] https://github.com/kata-containers/kata-containers/issues/149 [3] https://docs.google.com/document/d/1E3GLCzNgrcigUlgWAZYlgqNTdVwiMwCRTJ0QnJhL...
On 2 Jun 2021, at 09:52, Samuel Ortiz sameo@linux.intel.com wrote:
All,
During our May 27th meeting, we continued dicussing the image service offload proposal, based on a more detailed description of the overall proposed architecture [1].
We had a large crowd attending the meeting, including key CRIO and containerd maintainers. Eventually, there was agreement from both of these projects that moving the image pulling layer in the guest makes sense for the confidential computing use case we're trying to build here. I'm expecting the discussions on the implementation details and steps to continue happening through the relevant issue [2], but we now know that the potential impact on CRIO and containerd is acceptable.
Tomorrow, I'm proposing that we discuss about the Host/Guest API and how/if we'd want to split it. The agenda is at the usual place [3].
During the AC yesterday, Eric expressed some interest in the discussion, which he described as "not 4AM interest", so he asked if we could wait until next week to be in a US-compatible time slot.
Do we have something to fill the discussion this week? Or do we expect the API split discussion to last for two weeks (it might)?
Last but not least: Starting from tomorrow, this meeting will alternate between 2 time slots: 12 noon (UTC) and 3pm (UTC). Tomorrow's (June 3rd) meeting starts at noon UTC.
Cheers, Samuel.
[1] https://docs.google.com/presentation/d/1469nSRFtlHMSDDDWVLj0i21dR9M_3SO76ehQ... [2] https://github.com/kata-containers/kata-containers/issues/149 [3] https://docs.google.com/document/d/1E3GLCzNgrcigUlgWAZYlgqNTdVwiMwCRTJ0QnJhL...
kata-dev mailing list kata-dev@lists.katacontainers.io http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev
Hi Christophe,
See below:
On Wed, Jun 02, 2021 at 10:09:32AM +0200, Christophe de Dinechin wrote:
On 2 Jun 2021, at 09:52, Samuel Ortiz sameo@linux.intel.com wrote: Tomorrow, I'm proposing that we discuss about the Host/Guest API and how/if we'd want to split it. The agenda is at the usual place [3].
During the AC yesterday, Eric expressed some interest in the discussion, which he described as "not 4AM interest",
Right, there's not much that fits the 4am interest requirement :)
so he asked if we could wait until next week to be in a US-compatible time slot.
Do we have something to fill the discussion this week? Or do we expect the API split discussion to last for two weeks (it might)?
If we start the discussion this week, we must spend a fair chunk of next week's meeting summarizing what we'd discuss tomorrow to allow for folks to catch up. I'm fine with that as long as Eric and others are ok as well. If that does not work, tomorrow we can start discussing about attestation (local, remote, external APIs, etc).
Eric, please let us know what would work best for you.
Cheers, Samuel.
On 2 Jun 2021, at 10:35, Samuel Ortiz sameo@linux.intel.com wrote:
Hi Christophe,
See below:
On Wed, Jun 02, 2021 at 10:09:32AM +0200, Christophe de Dinechin wrote:
On 2 Jun 2021, at 09:52, Samuel Ortiz sameo@linux.intel.com wrote: Tomorrow, I'm proposing that we discuss about the Host/Guest API and how/if we'd want to split it. The agenda is at the usual place [3].
During the AC yesterday, Eric expressed some interest in the discussion, which he described as "not 4AM interest",
Right, there's not much that fits the 4am interest requirement :)
so he asked if we could wait until next week to be in a US-compatible time slot.
Do we have something to fill the discussion this week? Or do we expect the API split discussion to last for two weeks (it might)?
If we start the discussion this week, we must spend a fair chunk of next week's meeting summarizing what we'd discuss tomorrow to allow for folks to catch up. I'm fine with that as long as Eric and others are ok as well. If that does not work, tomorrow we can start discussing about attestation (local, remote, external APIs, etc).
Good idea. That's a great topic too.
Eric, please let us know what would work best for you.
Cheers, Samuel.
This works for me! Thanks ya'll.
On Wed, Jun 2, 2021 at 1:39 AM Christophe de Dinechin cdupontd@redhat.com wrote:
On 2 Jun 2021, at 10:35, Samuel Ortiz sameo@linux.intel.com wrote:
Hi Christophe,
See below:
On Wed, Jun 02, 2021 at 10:09:32AM +0200, Christophe de Dinechin wrote:
On 2 Jun 2021, at 09:52, Samuel Ortiz sameo@linux.intel.com wrote: Tomorrow, I'm proposing that we discuss about the Host/Guest API and how/if we'd want to split it. The agenda is at the usual place [3].
During the AC yesterday, Eric expressed some interest in the discussion, which he described as "not 4AM interest",
Right, there's not much that fits the 4am interest requirement :)
so he asked if we could wait until next week to be in a US-compatible time slot.
Do we have something to fill the discussion this week? Or do we expect the API split discussion to last for two weeks (it
might)?
If we start the discussion this week, we must spend a fair chunk of next week's meeting summarizing what we'd discuss tomorrow to allow for folks to catch up. I'm fine with that as long as Eric and others are ok as
well.
If that does not work, tomorrow we can start discussing about attestation (local, remote, external APIs, etc).
Good idea. That's a great topic too.
Eric, please let us know what would work best for you.
Cheers, Samuel.
kata-dev mailing list kata-dev@lists.katacontainers.io http://lists.katacontainers.io/cgi-bin/mailman/listinfo/kata-dev
participants (3)
-
Christophe de Dinechin
-
Eric Ernst
-
Samuel Ortiz