Today we’re releasing the R2 of the reference implementation of the Sovereign Cloud Stack. Our release announcement covers the many aspects of this release in detail. These range from our work on integrating the Kubernetes Cluster API, many improvements that focus on the operator experience up to items that help to assure standard conformance of clouds being built with SCS.
Since SCS has a fixed release schedule - we release twice a year - and through all the lively discussions of the past weeks moving towards our R2 we already have ideas of things we want to cover in the R3 cycle.
SCS builds on top of the interaction of community members - CSPs, operators of cloud infrastructure as well as integrators who either build or support the process of building cloud infrastructure.
There is a proposal for a gitops style cluster management approach on the table.
As part of our discussions in the last week around octavia we concluded that it makes sense to further invest into the topic of loadbalancing withing cloud infrastructure and see if the current status, not only concerning OpenStack, can be improved.
In R2 we started to lay the groundwork for the federation. Within the R3 cycle we want to to document and validate (and probably enhance) the user federation capabilities which we have implemented with keycloak and OpenID Connect federation.
One of the goals of SCS is to help operators gain more velocity with their cloud infrastructure. However this relies on intensive testing and excellent quality assurance. As part of this effort we have started to use zuul as CI framework and intend to make significant progress in connecting all the existing tests to it.
Good and extensive testing, while being important, are not the only aspects of providing solid cloud infrastructure. We want to heavily invest in SCS conformance testing, allowing providers that differ significantly from our reference implementation to still adjust their platforms for full compatibility. With the work that is happening in the Special Interest Group (SIG) Monitoring we want to make sure operators have an in-depth view into their stack. This will range from reworking the OpenStack Health Monitor to build upon existing components such as tempest and rally as well as making sure logs are aggregated and can be analyzed.
In the last months metering and billing were discussed within our community. It has become apparent that this is one of the topics where the SCS project can provide added value to the broader OpenStack community and this will be one of the items that will be followed upon in the next few months.
The scs community started to work on a disk encryption feature based on tang. We’ve seen first previews of it and a lightning talk covered this first preview. For R3 we intend to complete the disk encryption feature.
As with the last release we will dive into a short retrospective mode in the upcoming week. In our team sessions in the calendar week 13 we will focus on reviewing our collaboration of the last months and see what can be kept, improved, started or stopped. While there are ideas out on what R3 should bring our community is invited to join the discussion and help to shape SCS so that it adds value to your cloud infrastructure. The team meetings in week 14 are dedicated to these topics.
We invite our community to join the discussions in the corresponding team sessions.