We’re just one week ahead of the OpenInfra Summit 2022 and we’re thrilled to meet you all on-site again. This post is to inform you about our activities at this truly important and exciting event for our community. Numerous members will be present with various talks and sessions throughout all three days. We have collected all contributions of our community below1 and we hope to see you at one or the other presentation.
Schedule
10:10 – 10:17
The Intersection of Open Source Technologies and Digital Sovereignty
Digital sovereignty is a key priority for the German Government and the Ministry for Economic Affairs and Climate Action (BMWK). In her keynote, Franziska Brantner will elaborate on its strategic and political relevance and will shed light on the concrete activities of the BMWK to strengthen sovereignty with a special focus on open source. A followup Q&A with Frederic Lardinois will shed more light on the role open source plays in tackling digital sovereignty.
14:50 – 15:20
Digital sovereignty: Why open infrastructure matters
16:00 – 16:40
Fearless automation: Deploying OpenStack on top of Kubernetes with YAOOK, How a retailer became a Cloud Provider
STACKIT is part of the IT organization within the Schwarz Group, the largest European retail company. This includes Lidl and Kaufland, Schwarz Produktion in food production and PreZero in environmental services. In cooperation with Cloud and Heat the Schwarz Group created the Lifecycle Management Tool for OpenStack „Yaook“ (Yet another openstack on kubernetes). Since 2021 it is available as opensource. The Yaook project aims to be the next step in automating deployment and operation of OpenStack clouds. In its core, it relies on Kubernetes operators, small pieces of code which compare observed and intended state in order to determine actions to converge the system. This idea can, however, be taken to a next level: What if there was a control loop which managed an entire OpenStack deployment? Yaook provides just that. Beyond the initial installation, it supports „Day 2 operation“ tasks, such as upgrades and node maintenance / reinstallation. At the same time, Yaook knows its limits and will wait for human operator input in cases of potential loss of data or availability. That way, Yaook makes for a resilient and reliable life-cycle management of OpenStack clouds.
11:20 – 11:35
Digital sovereignty: Why open infrastructure matters
Hardly any other term has been redefined and reframed so frequently in public discussions in recent years as “digital sovereignty”. Interpretations range from using Open Source Software down to local franchise agreements with proprietary cloud providers applying the term as a boilerplate. This talk will demystify the term “digital sovereignty” and outline why it is not just enough to be sovereign on a higher container layer or create local spin-offs from proprietary cloud service providers. Together we will look at different criteria for digital sovereignty and realize that closed-source “sovereign” cloud offerings may not bring as much sovereignty as promised. Open infrastructure matters and is key to independence, trust and innovation. In order to achieve true digital sovereignty, we need open source, operating knowledge and vital open communities.
14:15 – 14:30
How to create beautiful cloud-native landscapes?
14:50 – 15:20
An Open Operations concept
16:00 – 16:30
Forum on Loadbalancer-as-a-Service - fostering Octavia and taking LBaaS to a next Level
Open Infrastructure technology has become increasingly mature. Nevertheless, high quality management and operations of open platforms continues to require DevOps teams with significant skills. The Sovereign Cloud Stack project has been created with the goal to significantly lower the bar. While the integration of a standardized operational stack helps the platform operations teams, it is only an important first step. The next step is to create strong communities that share operational practices and make them openly available. Documenting best practices, publishing the little tools to do cleanup jobs or the setup of dashboards that can be used for capacity management are all important contributions. Being transparent about bugs, incidents, security responses may sound disturbing to traditional Ops teams at first — yet it is key to build up experience. It also helps to create trust with users, especially when incidents do occur. This is the core of the Open Operations concept.
16:40 – 17:10
Intent-based holistic data center management
OpenStack is great for delivering large amounts of resources via APIs „as a service“. However, when it comes to enrolling and managing physical resources themselves, e.g. Switches or PDUs, or even more generally the entire data center, OpenStack itself reaches its limits. Is the IPAM the source of truth or OpenStack itself? How is the reconciliation done? Who approves changes there? How can annomalies be found? An intent-based holistic data center management approach approach allows a high abstraction of the underlying configurations and control. For example, it is thus possible to populate a pool of physical servers first with OpenStack. If the requirement changes in the future, a part can be removed from this pool of resources and populated with a completely independent workload, for example a Gardener on Metal. In this talk we will present our open source based approach (that is available as open source) with Netbox, Git, Python, Ansible and OpenStack Ironic in detail.
17:20 – 17:40
Sovereign OpenStack
17:20 – 17:40
Nova Compute REST API History
Have you ever wondered why the first version of Nova’s API is called v2? And what is the point of microversions anyway? What happened to that v3 Compute API? Why don’t you have any Beta APIs? What happened to the EC2 API? Why are those folks from Rackspace so touchy about HTTP 500 errors? How were all these decisions made? The history of the Nova API is really a history of the community discussions and people that shaped its direction. Back when I first started working with OpenStack Bexar and Citrix XenServer, things were somewhat different to today. We used euca2ools to start VMs. At Rackspace public cloud I spent some time worrying about HTTP 500 errors. At StackHPC I have spent time working with many different people using OpenStack APIs and they are frequently interested in the history of the OpenStack Compute API. Lets take a deep look at the history of a few API quirks and lessons I learnt about Open Development.
19:00 – ...?
Community Meetup
09:40 – 10:10
Open Operations: the Fifth Open paradigm
11:50 – 12:20
Observability in OpenStack: - Learnings and building blocks from the SIG Monitoring at SCS
14:40 – 15:10
Deprivileging of Service Accounts Between Individual OpenStack Services
14:40 – 15:50
The Future of the Public Cloud SIG
Community Meetup
We’re excited to invite you to our community gathering on Wednesday, June 8 (formerly known as „the secret party“). Let’s meet from 7 pm at the c-base e.V. (Rungestraße 20, 10179 Berlin), which is about 15 min by public transport from the Berlin Congress center (if you are probably attending the OpenInfra Summit).
Please note that you need a valid proof of vaccination or recovery to enter the c-base. Wearing a mask during the evening is optional.
Images by MeTaMiND EvoLuTioN MeTaVoLuTioN, CC BY 2.0, via Wikimedia Commons
Join us for food and drinks, for continuing the conversations started during the day or making plans for the next meeting. We thank our sponsors OSISM GmbH und Aitus UG for the food – they are both members of the Open Source Business Alliance e.V. – as well as the c-base for the great opportunity to have our social event in their space station.
