Day 2 at the KubeCon surpassed Day 1, and visitors, busy exploring the venue, were a little bit overwhelmed by the scope of event. Today, everyone is chill and relaxed: what a perfect time to start paying attention to knowledge! Everything at KubeCon, ranging from morning keynotes to booth crawls and lectures, serves the ultimate purpose of learning and building a vibrant community to promote cloud-native applications.
The beauty of KubeCon is that anyone — no matter what experience they have had with Kubernetes — can get up to speed pretty quickly with this technology. KubeCon sessions are tailored to different skill levels (Beginner, Intermediate, Advanced) and cover a wide array of topics, even exotic ones. Neither Kubernetes newbies or experts will feel left out while at KubeCon.
If you are a newbie to Kubernetes, you can check out some “intro” sessions. These provide with a powerful launchpad into the world of Kubernetes and cloud native technology.
There are a lot of sessions for Intermediate, and Advanced students of technology, too. The more advanced sessions will typically include some coding practices and walkthroughs, so don’t forget to take your laptop with you.
In general, you can choose among several session formats at KubeCon:
The topic coverage at KubeCon is tremendous, too. There are sessions on serverless technology, storage, container runtimes, Machine learning, CI/CD, and many more. It’s easy to miss something!
All projects hosted by CNCF are open source, so they welcome contributions from the community. KubeCon is definitely a great place to find your favorite project.
If you want to contribute to a project, you can meet its maintainers or representatives of the Special Interest Group (SIG) who oversee a specific part of the project. There are several options for you here:
KubeCon is a place where cutting-edge technologies are announced and effective industry solutions are shared. Today, we have picked several topics and sessions that we have found to be the most exciting.
One of the hottest topics in the cloud-native community today is debugging software and transactions in distributed environments such as clusters and cloud networks. The standards for such debugging are developed by the OpenTracing project, and it is well-represented at the KubeCon this year.
Ted Young, the Director of Open Source Development at LightStep and one of the core maintainers of the OpenTracing project, spoke about a new paradigm of Trace Driven Development that is closely related to the distributed tracing mentioned above. In brief, he proposed a new approach to code testing which he calls trace testing. The basic idea of trace testing is that we trace against trace data rather than code base. As it turns out, such an approach has a lot of advantages over traditional unit/integration test dichotomy. At the same time, trace testing retains unit-test-like ability of traditional tracing combining the best of two worlds. This synthesis makes the new approach suitable for microservices and serverless environments.
To be fast and efficient, Machine Learning requires scalable and distributed infrastructure. Kubernetes is a perfect solution for this. A recent marriage between Kubernetes and ML was heralded by the Kubeflow project dedicated to making deployments of ML workflows on Kubernetes simple, portable and scalable. KubeFlow is only one part of the immense efforts made by the community to adjust ML to Kubernetes. Two lecturers are worth our attention in this respect.
In their lecture titled Natural Language Code Search For GitHub Using Kubeflow Hamel Husain (Senior Data Scientist, GitHub) and Jeremy Lewi (Software Engineer, Google) demonstrated how Kubernetes and ML can be used together. They presented a search engine for GitHub code that uses natural language processing to find relevant code segments.
Huamin Chen (Principal Software Engineer, Red Hat) and Yuan Zhou (Senior Software Development Engineer, Intel) in their talk Nezha: a Kubernetes native Big Data Accelerator for Machine Learning introduced the audience to a new open source project called Nezha, which makes storing Big Data in Kubernetes more efficient. The motivation for this in-cluster storage solution is the high latency of AWS S3 and other cloud storages used for Big Data. The speakers decided to export Big Data to local caches inside Kubernetes clusters to make it directly available to ML tasks running in the cluster. This made these the ML tasks running on Kubernetes 2 times faster!
Another topic that drew our attention today is the Operator Framework for Kubernetes. The concept of software operators was introduced by CoreOS back in 2016. In a nutshell, an operator is any application-specific or domain-specific controller that extends the Kubernetes API to simplify deployment, configuration, and management of complex stateful applications on behalf of Kubernetes users. (We discussed how to use Operators in our tutorial about deploying Prometheus with the Prometheus Operator.)
A good introduction to the concept of Operators was made by Diane Mueller, Director, Community Development, Red Hat) and Sebastian Pahl (Director of Engineering at Red Hat) in this speech.
Closely related to this topic was the speech titled “Using Services outside of Kubernetes with Service Catalogue” delivered by Ara Pulido from Bitnami. Service catalogs allow users to easily combine Kubernetes with external services available from public cloud providers in a powerful way. With service catalogs, developers can easily delegate things like remote database management to public cloud-managed services while focusing on deploying their applications. Service catalog is definitely a good solution for building complex micro-services architectures and apps that span across different environments and rely on different cloud services.
KubeCon is truly a social place where people have fun and can explore the place/city they visit. Wednesday night KubeCon visitors had an opportunity to join an All Attendee Party sponsored by MongoDB and Tigera. The agenda included some of the most famous Seattle’s sites like Museum of Pop Culture (MoPoP), Chihuly Garden and Glass, and the Space Needle. Seattle’s iconic view from the Space Needle was really unforgettable!
We will be glad to see you by our booth on Day 3 at KubeCon and show you a demo of our new Supergiant 2.0.0 toolkit. We’re developing a great user community around the Supergiant toolkit and have a ton of new features to share with you. We will be glad to hear your feedback about the toolkit, too.
And don’t forget to ask for a Supergiant branded t-shirt and a Kubernetes dictionary with the most important cloud-native terms that help you navigate the complex cloud-native landscape. If you can’t make it to KubeCon, you can download the PDF version of our dictionary for free here. Enjoy!