My day 1 at KubeCon Europe 2022 started early and my expectations are high. I arrived after 7.30am and I was impressed how easy and fast the check-in was even I had to show my ID and COVID certificate, which would allow me to get inside to the badge printer.
The event location is quite huge and it is very important that you check the map before you enter the building. I had to learn it the hard way and had no clue where I find the breakout rooms or the event center, which was even a different building. The helpdesks were busy with a long queue of people who probably also were running around like me, like a headless chicken. đ But after a half day of running from one room to another, I found my way to the sessions.
65% of the KubeCon and CloudNativeCon visitors are first-timers like me!
The number if CNCF projects is growing fast
In 2016, the first KubeCon in North America had around 700 people
KubeCon 2022 Europe has >7’000 people onsite and >10’000 joining virtually
Priyanka’s message was about collaboration for long-term success. The CNCF is about a large continuously growing community that must work together, and each person can contribute in her/his own way.
These were the four big announcements she presented during the keynote session:
KubeCon Europe 2023 is going to happen in Amsterdam
After that Boeing presented their company and mission and we heard Mercedes Benz telling their story of “7 years running Kubernetes”. Mercedes Benz operates more than 900 Kubernetes clusters and 3’500 machines all over the world. They also mentioned that they migrated their clusters to Cluster API, a Kubernetes project with VMware as one of the main contributors.
After a short break it was time for my scheduled breakout sessions.
West Side CD: The Deployment Ballet Goes On
Benoit Moussaud from VMware Tanzu presented a different way of CI/CD with Cartographer, which ” is a Kubernetes-native Choreographer providing higher modularity and scalability for the software supply chain”:
Overview and State of Knative
The next session was in the event center building, which hosted the presentation of Mauricio Salatino, VMware and Carlos Santana, IBM. They gave an overview of the Knative philosophy of being “Kubernetes native”.
Knative offers a simplified developer experience deploying and managing stateless and event-driven applications. Maurico mentioned the following Knative features:
Simpler Abstractions
Autoscaling
Progressive Rollout
Event Integrations
Event Handling
Pluggable
From Kubernetes to PaaS to ⌠Err, Whatâs Next?
The third session I would like to highlight was from Daniel Bryant, Ambassador Labs. The key message of his presentation was about the “golden path” aka a paved platform. Â
At the beginning of his presentation Daniel started with the “real question” how much you should build yourself and how you should you assemble the control plane for effective use. Before going deeper into that topic he also joked around, and we know that 50% of all jokes are true, that the CNCF landscape and each KubeCon is not helping very much to make things for developers and operators easy:
It was very interesting and important to hear from him, that you cannot provide a good developer experience without a good user experience. Enterprises and platform teams need to treat the platform as a product and focus on tooling and interoperability.
Daniel said, that you need to think and design in/for different personas, user research is key and that you should watch your users doing things or using tools. Then you understand how you need can provide a good user experience for platforms. I personally believe that design thinking is key here.
Another interesting fact he mentioned is how the focus changed from the past KubeCons to the actual one. People focused mostly on operations and realized that the platform as a product mindset and approach is the way forward to provide also a good developer experience.
The VMware Tanzu Labs platform-as-product approach combines Product Management (PM), User-Centered Design (UCD), Agile, eXtreme Programming (XP), and Site Reliability Engineering (SRE) practices. A dedicated, balanced platform team uses these practices to both build and run the platform product.
Conclusion
It was a long day with a lot of impressions and new information. đ I definitely felt the spirit and the expertise of this large community! I am already excited and curious about tomorrow! Enjoy KubeCon Europe 2022!
PS: Are you looking for a job? KubeCon and the community got you covered!
Twitter and LinkedIn are flooded with posts about the KubeCon + CloudNativeCon Europe 2022 event happening in Valencia, Spain. It is one of the biggest conferences in the world, where users, developers, geeks and newbies come together, who are interested in cloud native standards, technologies and projects like Envoy, Fluentd, Harbor, Helm, Kubernetes, Open Policy Agent, Argo, Buildpacks, Cilium, Contour, Flux, SPIFFE, SPIRE and many more.
While May 16-17th consist of a pre-event program, which I don’t attend, I look forward to main conference from 18-20 May. For those who cannot or still don’t want to attend in person, KubeCon Europe is a hybrid event that can be joined virtually as well.
Why do I join KubeCon Europe?
The first reason is the fact that this is my first KubeCon Europe. I want to experience the spirit of the open source community and learn new things. I know some of the open source projects VMware is contributing to but there are so many awesome projects and technologies that I would like to understand better.
The second reason is VMware Tanzu. VMware has a booth at KubeCon and a lot of Tanzu folks are there that I would like to meet. Some of them are presenting as well and I would like to listen to their words and learn how they explain things, which helps me then to better serve my customers.
Session Abstract: In this session, we’ll give attendees an overview of the Knative philosophy of being Kubernetes-native and working well with existing Kubernetes tools. Then we’ll provide a demo of FaaS using Knative and conclude with a roadmap for what’s next. Most importantly, we’ll provide information on how you can get involved either as a contributor or end-user who wants to give feedback on its future direction. With its recent donation to the CNCF at the incubating level, there’s never been a better time to get started with Knative.
Session Abstract: This talk will look back on my experience of building platforms, both as an end-user and now as part of an organization helping our customers do the same. The key takeaways are: – Treat platform as a product – Realize that you canât have good developer experience (DevEx) without good UX – Focus on workflows and tooling interoperability Weâll wrap this talk with a walk-through of the CNCF ecosystem through the developer control plane lens, and look at whatâs next in the future of this important emerging category.
Session Abstract: Project Harbor is an open-source trusted cloud native registry project that stores, manages, signs, and scans content, thus resolving common image or Helm Chart management challenges. It has been widely used by organizations large and small around the world to resolve both the container image and Helm Chart management challenges. In this presentation, we will cover some advanced features of using Harbor, such as image signature management(cosign), image management in a cloud environment, unified management of Helm chart and container images, and highly-available deployments.Furthermore, the team would love to get feedback from users and contributors to current features and future roadmap.
Session Abstract: SPIFFE aims to strengthen the identification of software components in a common way that can be leveraged across distributed systems by anyone, anywhere. The ability to maintain software security by standardizing how systems define, attest, and maintain software identity, regardless of where systems are deployed or who deploys those systems, confers many benefits. The use of SPIFFE can significantly reduce costs associated with the overhead of managing and issuing cryptographic identity documents and accelerate development by removing the need for developers to understand the complexity involved to secure service-to-service communication, but that is not the only outcome. Production identity can have a positive impact on many areas such as interoperability, compliance, audibility, and more. This presentation demonstrates the real world scenarios and outcomes of deploying SPIFFE across your infrastructure and also using it to bridge and integrate the infrastructure of others.
Session Abstract: Cloud Provider code allows Kubernetes to run on top of different platforms, with an implementation for each. The agenda will include: An overall status report on removing the cloud provider code from the main Kubernetes repository to âout of tree repositories; âLightning talksâ for individual cloud providers, reporting efforts, accomplishments, and roadmap for features and getting “out-of-tree”. Weâll also discuss the plans to handle cloud provider migration – including interesting topics like building and migrating to cloud controller managers, and kubelet image credential providers. The goal of SIG Cloud Provider is to promote a vendor-neutral ecosystem for our community. We will close with details on how you can get involved with the SIG as either a cloud infrastructure supporter, a K8s distribution author, or a K8s user.
Session Abstract: The Cluster Lifecycle SIG is the Special Interest Group that is responsible for building the user experience for deploying and upgrading Kubernetes clusters. Our mission is examining how we should change Kubernetes to make it easier to operate. In this deep dive, we will examine how Cluster API simplifies the cluster management experience for cluster operators by enabling consistent machine management across environments and quick stamping of Clusters using some new exciting features like ClusterClass.
Session Abstract: Kubernetes abstracts out differences across hosting infrastructure, but there are cases when integrated monitoring across the layers of storage, compute, etc, are essential. When faults or reconfiguration happen, manual monitoring, diagnosis and remediation can be slow, costly, and error prone. The VMware Event Broker Appliance is an open-source project, usable with Cloud Events and Knative to optimize availability, auditing, compliance, etc. based on vSphere events. We’ll cover popular use cases and how to get started. The K8s VMware User Group shares best practices for hosting K8s on VMware infrastructure, and we will close the session with details on how you can participate in the group.
I’ll do my best to provide daily summaries of the sessions I attended.
See You in Valencia
I expect to learn a ton of new things and I hope to meet a lot of new faces.
If you are at KubeCon Europe and want to have a chat, you will most likely find me near the VMware Tanzu booth. đ
According to Gartner, regulated industry customers (such as finance and healthcare) and governments are looking for digital borders. Companies in these sectors are looking to reduce vendor lock-in and single points of failure with their cloud providers, whose data centers sometimes are also outside their country (e.g., Switzerland based customer with an AWS data center in Frankfurt).
The market for cloud technology and services is currently dominated by US and Asian cloud providers and many (European) companies store their data in these regions. There are European regions and data centers, but the geopolitical and legal challenges, concerns about data control, industry compliance and sovereignty are driving the creation of new national clouds.
That is why Gartner sees sovereign clouds as one of the emerging technologies, which is currently at the start of the August 2021 published hype cycle:
As an example and first use case I would mention the Swiss federal administration, which doesnât see the need for an independent technical infrastructure under public law.
In June 2021 they published the statement that they notified the following cloud providers to become part of the federal administrationâs initial multi-cloud architecture:
Amazon Web Services (AWS)
IBM
Microsoft
Oracle
Alibaba
There are several reasons (pricing, market share, local data center availability) that led to this decision to build a multi-cloud architecture with these cloud providers. But it was interesting to read that the government did an assessment and concluded that no technical independent infrastructure is needed â no need for a local sovereign cloud.
This means that they want to keep their existing data centers to provide infrastructure and data sovereignty.
Interestingly, the Swiss confederation is exploring initiatives for secure and trustworthy data infrastructure for Europe and is examining participation in GAIA-X.
Use Case 2 – Current Sovereign Cloud Providers
There are other examples where organizations and governments saw the need for a sovereign cloud. Having a public cloud providerâs data center in the same country does not necessarily mean, that itâs a sovereign cloud per se. Hyperscale clouds often rely on non-domestic resources that maintain their data centers or provide customer support.
Governments and regulated industries say that you need domestic resources to provide a true sovereign cloud.
A good example here is the UK government, who has chosen the provider UKCloud, that delivers a consistent experience that spans the edge, private cloud and sovereign cloud.
Another VMware sovereign cloud provider is AUCloud, who provides IaaS to the Australian government, defense, defense industries and Critical National Industry (CNI) communities.
The third example I would like to highlight is Saudi Telecom Company (STC), that brings sovereign cloud services to Saudi Arabia.
What do UKCloud, AUCloud and STC have in common? They all joined the pretty new VMware Sovereign Cloud initiative and built their sovereign clouds based on VMware technology.
Use Case 3 – Cloud Act
Another motivation for a sovereign cloud could be the Cloud Act, which is a U.S. law that gives American authorities unrestricted access to the data of American IT cloud providers. It does not matter where the data is effectively stored. In the event of a criminal prosecution, the authorities have a free hand and do not even have to notify the data owners.
What does this mean for cloud users? Because of the Cloud Act, they cannot be sure whether when and to what extent their data or the data of their customers will be read by foreign authorities.
Use Case 4 â GAIA-X
Let me quote the official explanation of GAIA-X:
The architecture of Gaia-X is based on the principle of decentralization. Gaia-X is the result of many individual data owners (users) and technology players (providers) – all adopting a common standard of rules and control mechanisms â the Gaia-X standard.
Together, we are developing a new concept of data infrastructure ecosystem, based on the values of openness, transparency, sovereignty, and interoperability, to enable trust. What emerges is not a new cloud physical infrastructure, but a software federation system that can connect several cloud service providers and data owners together to ensure data exchange in a trusted environment and boost the creation of new common data spaces to create digital economy.
Gaia-X aims to mitigate Europeâs dependency on non-European providers and there seems to be no pre-defined architecture or preferred vendor when it comes to the underlying cloud platform GAIA-X sits on top.
While one would believe that a sovereign cloud is mandatory for GAIA-X, it looks more like a cloud-agnostic data exchange platform hosted by European providers and customers.
I am curious how providers build, operate and maintain a sovereign cloud stack based on open-source software.
How real is the need for Sovereign Cloud?
If a company or government wants to keep, extend, and maintain their own local data centers, this is still a valid option of course. But the above examples showed that the need for sovereign clouds exists and that the global interest seems to be growing.
What is the VMware Sovereign Cloud Initiative?
In October 2021 VMware announced their VMware Sovereign Cloud initiative where they partnering with cloud service providers to deliver a sovereign cloud infrastructure with cloud services on top to customers in regulated industries.
To become a so-called VMware Sovereign Cloud Provider, partners must go through an assessment and meet specific requirements (framework) to show their capability to provide a sovereign cloud infrastructure.
VMware defines a sovereign cloud as one that:
Protects and unlocks the value of critical data (e.g., national data, corporate data, and personal data) for both private and public sector organizations
Delivers a national capability for the digital economy
Secures data with audited security controls
Ensures compliance with data privacy laws
Improves control of data by providing both data residency and data sovereignty with full jurisdictional control
VMware aims to help regulated industry and government customers to execute their cloud strategies by connecting them to VMware Sovereign Cloud Providers (like UKCloud, AUcloud, STC, Tietoevry, ThinkOn or OVHcloud).
Sovereign Cloud Providers in Switzerland
Currently, there is no official VMware sovereign cloud provider in Switzerland. We have a few and strong VMware cloud provider partners as part of the VMware Cloud Provider Program (VCPP):
Let us come back to the use case 1 with the Swiss federal administration. They are building a multi-cloud and would have in Switzerland a potential number of at least 10 cloud service providers, which could become an official VMware Sovereign Cloud Provider.
There are other Swiss providers who are building a sovereign cloud based on open-source technologies like OpenStack.
Hyperscalers like Microsoft or Google need to partner with local providers if they want to build a sovereign cloud and deliver services.
VMware already has 4300+ partners with the strategic partnerships and the same technology stack in 120+ countries and some of them are already sovereign cloud providers as mentioned before.
What are the biggest challenges with a multi-cloud and a sovereign cloud infrastructure?
What do you think are the biggest challenges of an organization that builds a multi-cloud with different public cloud providers and sovereign clouds?
Let me list a few questions here:
How can I easily migrate my workloads to the public or sovereign cloud?
How long does it take to migrate my applications?
Which cloud is the right one for a specific workload?
Do I need to refactor some of my applications?
How can I consistently manage and operate 5 different public/sovereign cloud providers?
What if I one of my cloud providers is not strategic anymore? How can I build a cloud exit strategy?
How do I implement and maintain security?
What if I want to migrate workloads back from a public cloud to an on-premises (sovereign) cloud?
Which Kubernetes am I going to use in all these different clouds?
How do I manage and monitor all these different Kubernetes clusters, networking and security policies, create secure application communication between clouds and so on?
How do I control costs?
These are just a small number of questions, but I think it would take your organization or your cloud platform team a while to come up with a solution.
What is the VMware approach? Let me list some other articles of mine that help you to better understand the VMware multi-cloud approach:
Public cloud providers build local data centers and provide data residency. Sovereign clouds provide data sovereignty. Resident data may be accessed by a foreign authority while data sovereignty refers to data being subject to privacy laws and governance structures within the nation where that data is collected.
Controlling the location and access of data in the cloud has become an important task for CIOs and CISOs and I personally believe that sovereign clouds are not becoming important in 2 or 3 years, they are already very important and relevant, and we can expect a growth in this area in the next months.
My conclusion here is, that sovereign clouds and the public clouds are not competitors, they complement each other.
Intrinsic security is something we heard a lot in the past from VMware and it was mostly used to describe the strategy and capabilities behind the Carbon Black portfolio (EDR) that is complemented with the advanced threat prevention from NSX (NDR), that form together the VMware XDR vision.
I see similarities between intrinsic security and workout I am doing in the gym. My goal is to build more strength and power, and to become healthier in general. For additional muscle gain benefits and to be more time efficient, I have chosen compound exercises. I am not a fan of single muscle group exercises, which involve isolation exercises. Our body has a lot of joints for different movements, and I think it’s just natural if you use multiple muscle groups and joints during a specific exercise.
Therefore, when you perform compound exercises, you involve different muscles to complete the movement. This improves your intermuscular coordination of your muscles. In addition, as everyone would tell you, these exercises improve your core strength and they let your body become a single unit.
While doing weight training, it is very important to use the proper technique and equipment. Otherwise, the risk for injuries and vulnerabilities increases.
This is what intrinsic security means for me! And I think this is very much relevant to understand when talking about DevSecOps.
Understanding DevSecOps
For VMware, talking to developers and talking about DevOps started in 2019 when they presented VMware Tanzu the first time at VMworld. The ideas and innovation behind the name “Tanzu” should bring developers and IT operators closer together for collaboration.
DevOps is the combination of different practices, tools and philosophies that should help an organization to deliver applications and services at a higher pace. In the example above it would mean, that application developers and operations teams are not working isolated in silos anymore, they become one team, a single unit. But technology plays very important role to support the success of the new mindset and culture!
DevOps is about efficiency and the automation of manual tasks or processes. You want to become fast, flexible and efficient. When you put security in the center of this, then we start talking about DevSecOps. You want to know if one of your muscles or parts of the body become weak (defect) or vulnerable.
Depending on where you are right now on this application modernization journey, doing DevSecOps could mean a huge cultural and fundamental change to how you develop applications and do IT operations.
For me, DevSecOps is not about bringing security tools together from different teams and technologies. If DevOps and DevSecOps mean that you must change your mindset, then it is maybe also about time to consider the importance of new technology choices.
If DevSecOps means that you put security in the center of a DevOps- or container-centric environment, then security must become an intrinsic part of a modern application supply chain.
The VMware Tanzu portfolio has a lot of products and services to bring developers, operations and security teams together.
Where do we start? We need to “shift left” and this means we need to integrate security already early in the application lifecycle.
Code – Spring Framework
Before you can deliver an application to your customer, you need to develop it, you need to code. Application frameworks are a very effective approach for developing more secure and optimized applications.
Frameworks help to write code faster and more efficient. Not only does a framework can save your developers a lot of coding effort, but it also comes with pre-defined templates. They incorporate best practices and help you simplifying the overall application architecture.
Why is this important? To achieve better security or a more secure cloud native application, it makes sense to standardize and automate. Automation is key for security. Standardization makes it easier to understand or reuse code. You can write all the code yourself, but the chances are high that someone else did parts of your work already. Less variability reduces complexity and therefore enhances security.
There is the open-source Spring Framework for example, which uses Java as the underlying language (or .NET for Steeltoe). Both projects are managed by VMware and millions of developers use them.
What happens next? You would now run your continuous integration (CI) process (integration tests, unit tests) and then you are ready to package or build your application.
Build – Tanzu Build Service (TBS)
So, your code is now good for release. If you want to deploy your application to a Kubernetes environment, then you need a secure, portable and reproducible build that can be checked for security vulnerabilities, and you need an easy way to patch those vulnerabilities.
How are you going to build your container image where you application is going to be built into? A lot of customers and vendors have a dockerfile based approach.
TBS is constantly looking for changes in your source code and then automatically builds an image based on that. This means with TBS you don’t need any advanced knowledge of container packaging formats or know how to optimally construct a container creation script for a given programming language.
Tanzu Build Service knows all the images you have built and understands all the dependencies and components you have used. If something changes, your image is going to be rebuilt automatically and then stored in a registry of your choice. More about the registry in a second.
What happens if a vulnerability comes out and one of your libraries, operating systems or components is affected? TBS would patch this vulnerability and all the affected downstream container images would be updated automatically.
Imagine how happy your CISO would be about this way of building secure container images! đ
Build – Harbor
We have now pushed our container image to a container repository, a so-called registry. VMware uses Harbor (open-source cloud native registry by VMware, donated to the CNCF in 2018) as an enterprise-grade storage for container images. Additionally, Harbor provides static analysis of vulnerabilities in images through open-source projects like Trivy and Clair.
We have now developed our applications and stored our packaged images in our Harbor registry. What else do we need?
Build – VMware Application Catalog (VAC)
Developers are not going to build everything by themselves. Other services like databases or caching are needed to build the application as well and there are so many known and pre-packaged open-source software freely available online. This brings additional security risks and provides malicious actors to publish container images that contain vulnerabilities.
How can you mitigate this risk and reduce the chance for a critical application outage or breach?
In 2019, VMware acquired Bitnami, which delivers and maintains a catalog of 130+ pre-packaged and ready-to-use open-source application components, that are “continuously maintained and verifiably tested for use in production environments”.
Known as VMware Application Catalog (VAC, formerly also known as Tanzu Application Catalog), VAC as a SaaS offering provides your organization a customizable private collection of open-source software and services, that can automatically be placed in your private container image registry. In this case in your Harbor registry.
Example apps that are supported today:
Language Runtimes
Databases
App Components
Developer Tools
Business Apps
Nodejs
MySQL
Kafka
Artifactory
WordPress
Python
PostgreSQL
RabbitMQ
Jenkins
Drupal
Ruby
MariaDB
TensorFlow
Redmine
Magento
Java
MongoDB
ElasticSearch
Harbor
Moodle
How does it work?
There are two product features that I would like to highlight:
Build-time CVE scan reports for container images using Trivy
Build-time Antivirus scans for container images using ClamAV
Your application, built by Tanzu Build Service and VMware Application Catalog, is complete now, and stored in your Harbor registry. And since you use VAC, you also have your “marketplace” of applications, that is curated by a (security) team in your organization.Â
Note: Yes, VAC is a SaaS hosted application and you may have concerns because you are a public/federal customer. That’s no problem. Consider VAC as your trusted source where you can copy things from. There is no data stored in the public cloud nor does it run anything up there. Download your packages from this trusted repository over to you air gapped environment.
Run – Tanzu Kubernetes Grid (TKG)
Your application is ready to be deployed and the next step is in your pipeline is “continuous deployment“. We finally can deploy our applications to a Kubernetes cluster.
Tanzu Kubernetes Grid or TKG is VMware’s own consistent and conformant Kubernetes distribution that can run in any cloud. VMware’s strategy is about running the same Kubernetes dial tone across data centers and public cloud, which enables a consistent and secure experience for your developers.
TKG has a tight integration with vSphere called “vSphere with Tanzu”. Since TKG is an enterprise-ready Kubernetes for a multi-cloud infrastructure, it can run also in all major public clouds.
If consistent automation is important to you and you want to run Kubernetes in an air gapped environment, where there is no AWS, Azure or any other major public cloud provider, then a consistent Kubernetes version like TKG would add value to your infrastructure.
Manage/Operate – Tanzu Mission Control (TMC)
How do we manage these applications on any Kubernetes cluster (VMware TKG, Amazon EKS, Microsoft AKS, Google GKE), that can run in any cloud?
Some organizations started with TKG and others already started with managed Kubernetes offerings like EKS, AKS or GKE. That’s not a problem. The question here is how you deploy, manage, operate, and secure all these different clusters.
VMware’s solution for that is Tanzu Mission Control, which is also a SaaS-based tool hosted by VMware, that is the first offering I’m going to cover, that is part of a global Tanzu control plane. TMC is a solution that makes your multi-cloud and multi-cluster Kubernetes management much easier.
With TMC you’ll get:
Centralized Cluster Lifecycle Management. TMC enables automated provisioning and lifecycle management of TKG cluster across any cloud. It provides centralized provision, scaling, upgrading and deletion functions for your Kubernetes clusters. Tanzu Mission Control also allows you to attach any CNCF-conformant Kubernetes cluster (K8s on-prem, K8s in public cloud, TKG, EKS, AKS, GKE, OpenShift) to the platform for management, visibility, and analytic purposes. I would expect that we can use TMC in the future to lifecycle managed offerings like EKS, AKS or GKE.
Centralized Policy Management. TMC has a very powerful policy engine to apply consistent policies across clusters and clouds. You can create security, access, network, quota, registry, and custom policies (Open Policy Agent framework).
Identity and Access Management. Another important feature you don’t want to miss with DevSecOps in mind is centralized authentication and authorization, and identity federation from multiple sources like AD, LDAP and SAML. Make sure you give the right people or project teams the right access to the right resources.
Cluster Inspection. There are to inspection that you can run against your Kubernetes clusters. TMC leverages the built-in open-source project Sonobuoy that makes sure your cluster are configured in a conformant way with the Cloud Native Computing Foundation (CNCF) standards. Tanzu Mission Control provides CIS Benchmark inspection as another option.
Tanzu Mission Control integrates with other Tanzu products like Tanzu Observability and Tanzu Service Mesh, which I’m covering later.
Connect – Antrea
VMware Tanzu uses Antrea as the default container network interface (CNI) and Kubernetes NetworkPolicy to provide network connectivity and security for your pods. Antrea is an open-source project with active contributors from Intel, Nvidia/Mellanox and VMware, and it supports multiple operating systems and managed Kubernetes offerings like EKS, AKS or GKE!
Antrea uses Open vSwitch (OvS) as the networking data plane in every Kubernetes node. OvS is a high performance and programmable virtual switch that not only supports Linux, but also Windows. VMware is working on the achievement to reach feature parity between them, and they are even working on the support for ARM hosts in addition to x86 hosts.
Antrea creates overlay networks using VXLAN or Geneve for encapsulation and encrypts node-to-node communication if needed.
Connect & Secure – NSX Advanced Load Balancer
Ingress is a very important component of Kubernetes and let’s you configure how an application can or should be accessed. It is a set of routing rules that describe how traffic is routed to an application inside of a Kubernetes cluster. So, getting an application up and running is only the half side of the story. The application still needs a way for users to access it. If you would like to know more about “ingress”, I can recommend this short introduction video.
While a project like Contour is a great open-source project, VMware recommends Avi (aka NSX Advanced Load Balancer) provides much more enterprise-grade features like L4 load balancing, L7 ingress, security/WAF, GSLB and analytics. If stability, enterprise support, resiliency, automation, elasticity, and analytics are important to you, then Avi Enterprise, a true software-defined multi-cloud application delivery controller, is definitely the better fit.
Secure – Tanzu Service Mesh (TSM)
Let’s take a step back and recap what we have achieve until here. We have a standardized and automated application supply chain, with signed container images, that can be deployed in any conformant Kubernetes cluster. We can also access the application from outside and pod-to-pod communication, so that applications can talk to each other. So far so far good.
Is there maybe another way to stitch these services together or “offload” security from the containers? What if I have microservices or applications running in different clouds, that need to securely communicate with each other?
A lot of vendors including VMware realized that the network is the fabric that brings microservices together, which in the end form the application. With modernized or partially modernized apps, different Kubernetes offerings and a multi-cloud environment, we will find the reality of hybrid applications which sometimes run in multiple clouds.
This is the moment when you need to think about the connectivity and communication between your app’s microservices. Today, many Kubernetes users do that by implementing a service mesh and Istio is most probably the most used open-source project platform for that.
The thing with service mesh is, while everyone thinks it sounds great, that there are new challenges that service mesh brings by itself. The installation and configuration of Istio is not that easy and it takes time. Besides that, Istio is also typically tied to a single Kubernetes cluster and therefore Istio data plane – and organizations usually prefer to keep their Kubernetes clusters independent from each other. This leaves us with security and policies tied to a Kubernetes cluster or cloud vendor, which leaves us with silos.
Tanzu Service Mesh, built on VMware NSX, is an offering that delivers an enterprise-grade service mesh, built on top of a VMware-administrated Istio version.
The big difference and the value that comes with Tanzu Service Mesh (TSM) is its ability to support cross-cluster and cross-cloud use cases via Global Namespaces.
A Global Namespace is a unique concept in Tanzu Service Mesh and connects resources and workloads that form the application into a virtual unit. Each GNS is an isolated domain that provides automatic service discovery and manages the following functions that are port of it, no matter where they are located:
Identity. Each global namespace has its own certificate authority (CA) that provisions identities for the resources inside that global namespace
Discovery (DNS). The global namespace controls how one resource can locate another and provides a registry.
Connectivity. The global namespace defines how communication can be established between resources and how traffic within the global namespace and external to the global namespace is routed between resources.
Security. The global namespace manages security for its resources. In particular, the global namespace can enforce that all traffic between the resources is encrypted using Mutual Transport Layer Security authentication (mTLS).
Observability. Tanzu Service Mesh aggregates telemetry data, such as metrics for services, clusters, and nodes, inside the global namespace.
Monitor – Tanzu Observability (TO)
Another important part of DevSecOps with VMware Tanzu is observability. What happens if something goes wrong? What are you doing when an application is not working anymore as expected? How do you troubleshoot a distributed application, split in microservices, that potentially runs in multiple clouds?
Image an application split into different smaller services, that are running in a pod, which could be running in a virtual machine on a specific host in your on-premises datacenter, at the edge, or somewhere in the public cloud.
You need a tool that supports the architecture of a modern application. You need a solution that understands and visualizes cloud native applications.
That’s when VMware suggests Tanzu Observability to provide you observability and deep visibility across your DevSecOps environment.
Tanzu Observability has an integration with Tanzu Mission Control, which has the capability then to install the Wavefront Kubernetes collector on your Kubernetes clusters. The name “Wavefront” comes from the company Wavefront, which VMware acquired in 2017.
Since Tanzu Observability is only offered as a SaaS version, I would like to highlight that it is “secure by design” according to VMware:
Isolation of customer data
User & Service Account Authentication (SSO, LDAP, SAML)
RBAC & Authorization
Data encryption at rest and in transit
Data at rest is managed by AWS S3 (protected by KMS)
Certifications like ISO 27001/27017/27018 or SOC 2 Type 1
Summary – Tanzu Portfolio Capabilities
The container build and deploy process consists of the Spring runtime, Tanzu Application Catalog and Tanzu Build Service.
The global control plane (SaaS) is formed by Tanzu Mission Control, Tanzu Service Mesh and Tanzu Observability.
The networking layer consists of NSX Advanced Load Balancer for ingress & load balancing and uses Antrea for container networking.
The foundation of this architecture is built on VMware’s Kubernetes runtime called Tanzu Kubernetes Grid.
Another solution that might be of interest for you is Carbon Black Container. CB Container also provide visibility and control that DevSecOps team need to secure Kubernetes clusters and the application the deploy on top of them.
This solution provides container vulnerability & risk dashboard, image scanning, compliance policy scanning, CI/CD integration, integration with Harbor and supports any upstream Kubernetes like TKG, EKS, AKS, GKE or OpenShift.
Conclusion
DevSecOps with VMware Tanzu helps you to simplify and secure the whole container and application lifecycle. VMware has made some strategic acquisitions (Heptio, Pivotal, Bitnami, Wavefront, Octarine, Avi Networks, Carbon Black) in the past to become a major player the world of containerization, Kubernetes and application modernization.
I personally believe that VMware’s approach and Tanzu portfolio have a very strong position in the market. Their modular approach and the inclusion of open-source projects is a big differentiator. Tanzu is not just about Kubernetes, it’s about building, securing and managing the applications.
If you have a strong security focus, VMware can cover all the layers up from the hypervisor to the applications that can be deployed in any cloud. That’s the strength and unique value of VMware: A complete and diverse portfolio with products, that provide even more value when combined together.
Don’t forget, that VMware is number 1 when it comes to data center infrastructures and most of the customer workloads are still running on-premises. That’s why I believe that VMware and their Tanzu portfolio are very well positioned.
The customers I worked with last year were large enterprises with a multi-cloud strategy and they have just started their application modernization journey. Typically, VMware customers interested in Tanzu would take a look at the Standard edition first, which gives you:
Tanzu Kubernetes Grid Runtime
Tanzu Mission Control Standard
Avi Essentials (NSX Advanced Load Balancer)
Antrea (open-source) for container networking
and some other open-source software like Prometheus, Grafana, Fluent Bit, Contour
A lot of my customers were interested in Tanzu Advanced, but they were asking for something in between these editions. Tanzu Standard sounded very interesting, but almost all of them asked the followings questions:
What if I don’t build or modernize my own applications yet and get my application as a container from my ISV?
Prometheus and Grafana are nice, but I would like to have something more enterprise-ready for observability. How can I get Tanzu Observability?
Avi Essentials sounds great, but I am thinking to replace my current load balancer. Is it possible to replace my F5 or Citrix ADC (formerly known as Citrix NetScaler) appliances?
Contour seems to be a nice open-source project, but I am looking for something with built-in automation and analytics capabilities for ingress. Can’t I get Avi Enterprise for that as well?
I am looking for zero trust application security. How can you help me to encrypt traffic between containers or microservices, which could also be hosted on different clouds (e.g., on-prem and public cloud)?
The answer to these questions is Tanzu Kubernetes for Operations. Tanzu for Kubernetes Operations (TKO) is a bundle of VMware products and services to meet the requirements of cloud platform teams. It provides a centralized, consistent and simplified container management and operations across clouds and currently includes the following products and services:
Tanzu Standard Runtime (includes Tanzu Kubernetes Grid + open-source software from Tanzu Std), licensed per core
Antrea Advanced, licensed based on core (Antrea feature comparison can be found here)
NSX ALB (Avi) Enterprise, licensed based on service cores (not included, separate TKO SKU expected for end of Q1 2022)
Tanzu Activation Services for TKO deployment and configuration, includes migration of one upon-agreed application (not included in the TKO SKU, coming as an add-on)
Important Note: The VMware product guide says that “a Core is a single physical computational unit of the Processor which may be presented as one or more vCPUs“. So, if you plan a CPU overcommit of 1:2 (cores:vCPU) for your on-premises infrastructure, then you have to license 12 cores only.
Use this link to get additional information how to deploy and configure Tanzu Mission Control, Tanzu Observability and Tanzu Service Mesh.
What is Application Transformer for Tanzu?
Application Transformer for VMware Tanzu became generally available in February 2022.
Application Transformer can help you to convert virtual machines and application components to OCI-compliant container images, that then can be deployed into the Tanzu Kubernetes stack.
Tanzu App Navigator
Application Transformer helps you to analyze and visualize application components and dependencies. It also provides customers scores that allow them to decide which applications should be transformed.
App Navigator is a 4-to-6 week engagement that helps you to decide which applications you should tackle first and how much change is needed to drive business outcomes. It’s one thing to containerize an application, but App Navigator helps you to create a modernization strategy based on your goals.
Note: VMware’s App Navigator team uses Application Transformer during their service engagement.
Tanzu Application Platform
Deploying an application on Kubernetes is not an easy thing if you don’t know anything about Kubernetes.
If you would like to focus more on your applications and your developer’s experience, then Tanzu Application Platform (TAP) could be very interesting for you.
With Tanzu Application Platform, application developers and operations teams can build and deliver a better multi-cloud developer experience on any Kubernetes distribution, including Azure Kubernetes Service, Amazon Elastic Kubernetes Service, Google Kubernetes Engine, as well as software offerings like Tanzu Kubernetes Grid.
VMware is known to provide reduction of complexity and to provide cloud-agnostic infrastructures. They started to abstract the underlying server hardware, then the virtualization of the whole data center (compute, storage, network) came and the next step was the abstraction of public clouds like AWS, Azure and Google.
In the case of Tanzu Application Platform we are talking about an opinionated grouping of separate components that run on any conformant Kubernetes cluster (TKG, AKS, EKS, GKE, OpenShift etc.). From an application developer perspective an application can automatically be built, tested and deployed on Kubernetes.
Meaning, with TAP you get a modular application developer PaaS (adPaaS) offering and true application platform portability with the capability of “bring-your-own-Kubernetes”.
While I was working with one of the largest companies in the world during the past year, I learned a lot about VMware Tanzu and NSX Advanced Load Balancer (formerly known as Avi). Application modernization and the containerization of applications are very complex topics.
Customers are looking for ways to “free” their apps from infrastructure and want to go cloud-native by using/building microservices, containers and Kubernetes. VMware has a large portfolio to support you on your application modernization journey, which is the Tanzu portfolio. A lot of people still believe that Tanzu is a product – it’s not a product. Tanzu is more than just a Kubernetes runtime and as soon as people like me from VMware explain you the capabilities and possibilities of Tanzu, one tends to become overwhelmed at first.
Why? VMware’s mission is always to abstract things and make things easier for you but this doesn’t mean you can skip a lot of the questions and topics that should be discussed:
Where should your containers and microservices run?
With these kind of questions, you and I would figure out together, which Tanzu edition makes the most sense for you. Looking at the VMware Tanzu website, you’ll find four different Tanzu editions:
If you click on one of the editions, you get the possibility to compare them:
Based on the capabilities listed above, customers would like to know the differences between Tanzu Standard and Advanced. Believe me, there is a lot of information I can share with you to make your life easier and to understand the Tanzu portfolio better. đ
1) VMware Tanzu Standard and Advanced Features and Components
Let’s start looking at the different capabilities and components that come with Tanzu Standard and Advanced:
Tanzu Standard focuses very much on Kubernetes multi-cloud and multi-cluster management (Tanzu Kubernetes Grid with Tanzu Mission Control aka TMC), Tanzu Advanced adds a lot of capabilities to build your applications (Tanzu Application Catalog, Tanzu Build Service).
2) Tanzu Mission Control Standard and Advanced
Maybe you missed it in the screenshot before. Tanzu Standard comes with Tanzu Mission Control Standard, Tanzu Advanced is equipped with Tanzu Mission Control Advanced.
Note: Announced at VMworld 2021, there is now even a third edition called Tanzu Mission Control Essentials, that was specifically made for VMware Cloud offerings such as VMC on AWS.
I must mention here, that you could leverage the “free tier” of Tanzu Mission Control called TMC Starter. It can be combined with the Tanzu Community Edition (also free) for example or with existing clusters from other providers (AKS, GKE, EKS).
With Tanzu Mission Control Advanced you also get “CIS Benchmark inspections”
What if I want Tanzu Standard (Kubernetes runtime with Tanzu Mission Control and some open- source software) but not the complete feature set of Tanzu Mission Control Advanced? Let me answer that question a little bit later. đ
3) NSX Advanced Load Balancer Essentials vs. Enterprise (aka Avi Essentials vs. Enterprise)
Yes, there are also different NSX ALB editions included in Tanzu Standard and Advanced. The NSX ALB Essentials edition is not something that you can buy separately, and it’s only included in the Tanzu Standard edition.
The enterprise edition of NSX ALB is part of Tanzu Advanced but it can also be bought as a standalone product.
Here are the capabilities and differences between NSX ALB Essentials and Enterprise:
So, the Avi Enterprise edition provides a fully-featured version of NSX Advanced Load Balancer while Avi Essentials only provides L4 LB services for Tanzu.
Note: Customers can create as many NSX ALB / Avi Service Engines (SEs) as required with the Essentials edition and you still have the possibility to set up a 3-node NSX ALB controller cluster.
Important: It is not possible to mix the NSX ALB controllers from the Essentials and Enterprise edition. This means, that a customer, that has NSX ALB Essentials included in Tanzu Standard, and has another department using NSX ALB Enterprise for another use case, needs to run separate controller clusters. While the controllers don’t cost you anything, there is obviously some additional compute footprint coming with this constraint.
FYI, there is also a cloud-managed option for the Avi Controllers with Avi SaaS.
What if I want the complete feature set of NSX ALB Enterprise? Let’s put this question also aside for a moment.
4) Container Ingress with Contour vs. NSX ALB Enterprise
Ingress is a very important component of Kubernetes and let’s you configure how an application can or should be accessed. It is a set of routing rules that describe how traffic is routed to an application inside of a Kubernetes cluster. So, getting an application up and running is only the half side of the story. The application still needs a way for users to access it. If you would like to know more about “ingress”, I can recommend this short introduction video.
While Contour is a great open-source project, Avi provides much more enterprise-grade features like L4 LB, L7 ingress, security/WAF, GSLB and analytics. If stability, enterprise support, resiliency, automation, elasticity and analytics are important to you, then Avi Enterprise is definitely the better fit.
To keep it simple: If you are already thinking about NSX ALB Enterprise, then you could use it for K8s Ingress/LB and so much other use cases and services! đ Â
5) Observability with Grafana/Prometheus vs. Tanzu Observability
I recently wrote a blog about “modern application monitoring with VMware Tanzu and vRealize“. This article could give you a better understanding if you want to get started with open-source software or something like Tanzu Observability, which provides much more enterprise-grade features. Tanzu Observability is considered to be a fast-moving leader according to the GigaOm Cloud Observability Report.
What if I still want Tanzu Standard only but would like to have Tanzu Observability as well? Let’s park this question as well for another minute.
6) Open-Source Projects Support by VMware Tanzu
The Tanzu Standard edition comes with a lot of leading open-source technologies from the Kubernetes ecosystem. There is Harbor for container registry, Contour for ingress, Grafana and Prometheus for monitoring, Velero for backup and recovery, Fluentbit for logging, Antrea and Calico for container networking, Sonobuoy for conformance testing and Cluster API for cluster lifecycle management.
VMware is actively contributing to these open-source projects and still wants to give customers the flexibility and choice to use and integrate them wherever and whenever you see fit. But how are these open-source projects supported by VMware? To answer this , we can have a look at the Tanzu Toolkit (included in Tanzu Standard and Advanced):
Tanzu Toolkit includes enterprise-level support for Harbor, Velero, Contour, and Sonobuoy
Tanzu Toolkit provides advisoryâor best effortâguidance on Prometheus, Grafana, and Alertmanager for use with Tanzu Kubernetes Grid. Installation, upgrade, initial tooling configuration, and bug fixes are beyond the current scope of VMwareâs advisory support.
7) Tanzu Editions Licensing
There are two options how you can license your Tanzu deployments:
Per CPU Licensing – Mostly used for on-prem deployments or where standalone installations are planned (dedicated workload domain with VCF). Tanzu Standard is included in all the regular VMware Cloud Foundation editions.
Per Core Licensing – For non-standalone on-prem and public cloud deployments, you should license Tanzu Standard and Advanced based on number of cores used by the worker (including control plane VMs) and management nodes delivering K8s clusters. Constructs such as “vCPUs”, “virtual CPUs” and “virtual cores” are proxies (other names) for CPU cores.
Tanzu Advanced is sold as a “pack” of software and VMware Cloud service offerings. Each purchased pack of Tanzu Advanced equals 20 cores. Example of 1 pack:
Tanzu Observability: 160 PPS (sufficient to collect metrics for the infrastructure)
Tanzu Mission Control Advanced: 20 cores
Tanzu Service Mesh Advanced: 20 cores
NSX ALB Enterprise: 1 CPU = 1/4 Avi Service Core
Tanzu Standard Runtime: 20 cores
If you need more details about these subscription licenses, please consult the VMware Product Guide (starting from page 37).
As you can see, a lot of components (I didn’t even list all) form the Tanzu Advanced edition. The calculation, planning and sizing for the different components require multiple discussions with your Tanzu specialist from VMware.
8) Tanzu Standard Sizing
Disclaimer – This sizing is based on my current understanding, and it is always recommended to do a proper sizing with your Tanzu specialists / consultants.
So, we have learnt before that Tanzu Standard licensing is based on cores, which are “used by the worker and management nodes delivering K8s clusters”.
The three Supervisor Cluster (management nodes) VMs have each 4 vCPUs – this means in total 12 vCPUs.
The three Tanzu Kubernetes Cluster control plane VMs have each 2 vCPUs – this means in total 6 vCPUs.
The three Tanzu Kubernetes Cluster worker nodes (small size) have each 2 vCPUs – this means in total 6 vCPUs.
My conclusion here is that you need to license at least 24 vCPU to get started with Tanzu Standard.
Important Note: The VMware product guide says that “a Core is a single physical computational unit of the Processor which may be presented as one or more vCPUs“. If you are planning a CPU overcommit of 1:2 (cores:vCPU) for your on-premises infrastructure, then you have to license 12 cores only.
Caution: William Lam wrote about the possibility to deploy single or dual node Supervisor Cluster control plane VMs. It is technically possible to reduce the numbers of control plane VMs, but it is not officially supported by VMware. We need to wait until this feature becomes available in the future.
It would be very beneficial for customers with a lot of edge locations or smaller locations in general. If you can reduce the Supervisor Cluster down to two control plane VMs only, the initial deployment size would only need 14 vCPUs (cores).
9) NSX Advanced Load Balancer Sizing and Licensing
General licensing instructions for Avi aka NSX ALB (Enterprise) can be found here.Â
NSX ALB is licensed based on cores consumed by the Avi Service Engines. As already said before, you won’t be charged for the Avi Controllers and itt is possible to add new licenses to the ALB Controller at any time. Avi Enterprise licensing is based on so-called Service Cores. This means, one vCPU or core equals one Service Core.
Avi as a standalone product has only one edition, the fully-featured Enterprise edition. Depending on your needs and the features (LB, GSLB, WAF, analytics, K8s ingress, throughput, SSL TPS etc.) you use, you’ll calculate the necessary amount of Service Cores.
It is possible to calculate and assign more or less than 1 Service Core per Avi Service Engine:
25 Mbps throughput (bandwidth) = 0.4 Service Cores
200 Mbps throughput = 0.7 Service Cores
Example: A customer wants to deploy 10 Service Engines with 25MB and 4 Service Engines with 200MB. These numbers would map to 10*0.4 Service Cores + 4*0.7 Services Cores, which give us a total of 6.8 Service Cores. In this case you would by 7 Service Cores.Â
10) Tanzu for Kubernetes Operations (TKO)
Now it’s time to answer the questions we parked before:
What if I want Tanzu Standard (Kubernetes runtime with Tanzu Mission Control and some open- source software) but not the complete feature set of Tanzu Mission Control Advanced?
What if I want the complete feature set of NSX ALB Enterprise?
What if I still want Tanzu Standard only but would like to have Tanzu Observability as well?
Wherever you are on your application modernization journey, VMware and their Tanzu portfolio got your back covered. Not matter if you want to start small, make your first steps and experiences with open-source projects, or if you want to have a complete set with the Tanzu Advanced edition, VMware offers the right options and flexibility.
I hope my learnings from this customer engagement help you to better understand the Tanzu portfolio and its capabilities.
Please leave your comments and thoughts below. đ
Michael Rebmann is currently working as a Lead Solution Architect for VMware by Broadcom in Switzerland and focuses on some of the largest and most strategic customers. Michael engages with the community as a co-leader of the Swiss German VMUG and became a VMware vExpert PRO in 2022. Opinions are his own.
VMware CMTY Podcast #607 – VMware Cloud Foundation
VMware CMTY Podcast #546 – VMware Carbon Black Cloud Workload
Priyanka’s message was about collaboration for long-term success. The CNCF is about a large continuously growing community that must work together, and each person can contribute in her/his own way.
It was very interesting and important to hear from him, that you cannot provide a good developer experience without a good user experience. Enterprises and platform teams need to treat the platform as a product and focus on tooling and interoperability.
 
