Hi Ildiko,

If this helps, the latest TripleO architecture for edge computing is documented in our deployment guide:

https://docs.openstack.org/project-deploy-guide/tripleo-docs/latest/features/distributed_compute_node.html

https://docs.openstack.org/project-deploy-guide/tripleo-docs/latest/features/distributed_multibackend_storage.html

The architecture consists of a centralized control plane, and distributed sites for compute nodes (DCN) and if necessary, persistent storage with Ceph. Each site has an independent Ceph cluster serving only that site, as well as associated cinder-volume services plus HAProxy for some HA/failover. The storage configuration takes advantage of the Glance multi-store and associated management commands feature. The improvements around storage might be nice to highlight since I believe they were added specifically for the edge use case.

There is a downstream blog post done by Red Hat about storage at the edge, but it is entirely based on the upstream OpenStack TripleO, Glance and Cinder projects:

https://www.redhat.com/en/blog/introduction-openstacks-distributed-compute-nodes

https://www.redhat.com/en/blog/storage-and-distributed-compute-nodes-bringing-cinder-persistent-volumes-edge

There is also an earlier post about management of distributed sites with TripleO:

https://www.redhat.com/en/blog/multi-stack-deployments-edge-red-hat-openstack-platform-161

HTH, let me know if I can provide any further details.

On Mon, Feb 8, 2021 at 12:50 PM Ildiko Vancsa <ildiko.vancsa@gmail.com> wrote:

Hi,

I’m reaching out as I need a little guidance.

I was asked to give a presentation about the latest features and improvements in OpenStack that are relevant for edge computing. I wanted to ask for the community’s help, as the project teams are the experts, to give any pointers that you think would be great to include in my presentation.

I appreciate all your guidance in advance.

Thanks,
Ildikó