VMworld 2021 is going to happen from October 6-7, 2021 (EMEA). This year you can expect so many sessions and presentations about the options you have when combining different products together, that help you to reduce complexity, provide more automation and therefore create less overhead.
Let me share my 5 personal favorite picks and also 5 recommended sessions based on the conversations I had with multiple customers this year.
Project Monterey was announced in the VMworld 2020 keynote. There has been tremendous work done since then. Hear Niels Hagoort and Sudhansu Jain talking about SmartNICs and how they will redefine the data center with decoupled control and data planes – for ESXi hosts and bare-metal systems. They are going to cover and demo the overall architecture and use cases!
Learn from Matt Coppinger how augmented realited (AR) and virtual reality (VR) are transforming employee productivity, and how these solutions can be deployed and managed using VMware technologies. Matt is going to cover the top enterprise use cases for AR/VR as well as the challenges you might face deploying these emerging technologies. Are you interested how to architect and configure VMware technologies to deploy and manage the latest AR/VR technology, applications and content? If yes, then this session is also for you.
I am very interested to learn more cybersecurity. With Chad Skipper VMware has an expert who can give insights on how the Network Detection and Response (NDR) capabilities if NSX Advanced Threat Prevention provide visibility, detection and prevention of advanced threats.
Learn more about NUMA from Frank Denneman. You are going to learn more about the underlying configuration of a virtual machine and discover the connection between the Generapl-Purpose Graphics Processing Unit (GPGPU) and the NUMA node. You will also understand after how your knowledge of NUMA concepts in your cluster can help the developer by aligning the Kubernetes nodes to the physical infrastructure with the help of VM Service.
Are you interested to learn more about how to protect, detect, respond to and recover from cybersecurity attacks across all technology stacks, regardless of their purpose or location? Learn more from Amanda Blevins about the VMware solutions for end users, private clouds, public clouds and modern applications.
5 Recommended Sessions based on Customer Conversations
A lot of work is needed to better understand cryptographic agility and how we can address and manage the expected challenges that come with quantum computing. Hear VMware’s engineers from the Advanced Technology Group talking about the requirements of crypto agility and VMware’s recent research work on post-quantum cryptography in the VMware Unified Access Gateway (UAG) project.
Let Chris Wolf give you some insight into VMware’s strategic direction in support of edge computing. He is going to talk about solutions that will drive down costs while accelerating the velocity and agility in which new apps and services can be delivered to the edge.
In this session one can see how you can use two capabilities in VMware Tanzu Advanced, Tanzu Build Service and Tanzu Application Catalog, to feed a continuous stream of patched and compliant containers into your continuous delivery (CD) system. A must attend session delivered by David Zendzian, the VMware Tanzu Global Field CISO.
VMware NSX firewall reimagines East-West security by using a distributed- and software-based approach to attach security policies to every workload in any cloud. Chris Kruegel gives you insights on how to stop lateral movement with advanced threat prevention (ATP) capabilities via IDS/IPS, sandboxing, NTA and NDR.
Hear different the VMware CTOs Shawn Bass, Pere Monclus and Scott Lundgren talking about a zero trust approach. Shawn and the others will discuss specific capabilities that will enable customers to achieve a zero trust architecture that is aligned to the NIST guidance and covers secure access for users as well secure access to workloads.
For most organizations it is still new that they can talk about cybersecurity with VMware. VMware’s intrinsic security vision is something we have seen the first time at VMworld 2019, and since then it has become more a strategy than a vision.
VMware is not new to enterprise security and it didn’t start with Workspace ONE nor with NSX. Security was already part of their DNA since it was possible for the first time that two virtual machines can share a physical host and have isolated compute resources assigned.
Another example of (intrinsic) security came with vSAN and the encryption of data at rest, then followed by unified endpoint management and identity/access management with Workspace ONE. But wait!
It was August 2013 when Pat Gelsinger introduced NSX as the platform for network virtualization, which included the distributed firewall capability already. The internal firewall is built into the VMware hypervisor since almost 8 years now, wow!
NSX Service-Defined Firewall
I had no customer so far, who wasn’t talking about achieving zero trust security with micro-segmentation to prevent lateral (east-west) movement. Zero trust is one approach to improve data center defenses with the inspection of every traffic flow within the data center. The idea is to divide the data center infrastructure into smaller security zones and that the traffic between the zones is inspected based on the organization’s defined policies.
Micro-segmentation puts a firewall to each virtual machine or workload, allowing us to protect all east-west communication.
So, deploy micro-segmentation and the problem is solved, right? Not quite. While the concept of micro-segmentation has been around for a while, organizations still face barriers when trying to apply it in practice.
Let’s have a look at some of the barriers to micro-segmentation and why this solution alone is not enough (anymore) to achieve zero trust:
Policy discovery challenges – Identifying the right micro-segments and configuring the proper security policies is an extremely daunting task, especially in a dynamic data center environment.
Limited-access controls – Basing micro-segmentation solely on L4 attributes (e.g., IP addresses and ports) is not enough. The ephemeral nature of applications and flows requires more than that.
Reliance on agents – Some micro-segmentation implementations require the installation of extra software agents on each virtual machine (VM), causing complexity and introducing vulnerability.
Lack of threat detection and prevention – Threats often masquerade as normal-looking traffic. Settling for basic traffic blocking rules isn’t enough.
What does that tell us? Understanding the current applications’ topology and communication flows between their sub-services and -components is not easy. And with applications, which become less monolithic but very dynamic and distributed across multiple clouds, it becomes almost impossible, right?
NSX Intelligence is a home-grown solution that automates policy discovery, understands the communication between services and can construct apps and flows maps (topologies).
Can we assume that traffic from A to B over HTTPS is safe per se with micro-segmentation? Nope.
If we want to enhance traffic analysis capabilities and have a deeper look into traffic, the L7 (application layer) capabilities for micro-segmentation can be used.
Firewall rules cannot consume application IDs. A context-aware firewall identifies applications and enforces a micro-segmentation for east-west traffic, independent of the port that the application uses.
Other use case: For virtual desktop infrastructures (VDI), you could use VMware NSX’s ability to provide Active Directory identity-based firewall (IDFW) rules.
Okay. We have a topology now and can create context-aware service-defined firewall rules. How can we differentiate between good or bad traffic? How can we detect network anomalies?
Today’s attacks are becoming more sophisticated and hackers use masquerading techniques to embed threats within normal-looking traffic flows. Micro-segmentation alone will not intercept hidden threats, it only identifies traffic flows that should be allowed or blocked.
It’s time to talk about advanced inspection capabilities.
NSX Distributed IDS/IPS
In general, for a firewall to inspect traffic, the traffic has to pass through it. In a virtual world this means we would redirect traffic from the VM’s to the firewalls and back. A practice called hair-pinning:
That results in additional traffic and unnecessary latency. NSX has a distributed architecture, there is no centralized appliance that limits security capacity and network traffic doesn’t need to be hair-pinned to a network security stack for traffic inspection. Everything done with physical appliance can now be done in software (see coloring).
The term intrinsic security always means that security is built into the infrastructure. The micro-segmentation capabilities including NSX Intelligence come without an agent – no reliance on agents!
The VMware NSX Distributed IDS/IPS functionality adds additional traffic inspection capabilities to the service-defined firewall and follows the same intrinsic security principles.
Note: These regular-expression IDS/IPS engines detect traffic patterns and are programmed to look for malicious traffic patterns.
In my understanding, Lastline’s core product was a malware sandbox that can go deeper (than other sandboxes from other vendors) by using a full-system emulation to look at every instruction the malware executes.
The Lastline system uses machine learning that recognizes essential elements of an attack, unlike the narrow signature-based systems that miss the many variants an attacker may use. The Lastline approach is not just anomaly detection – anomaly detection treats every outlier as bad and results in many false positives. Lastline leverages the deep understanding of malicious behavior to flag clearly bad activities such as East-West movement, command and control activity, and data exfiltration.
This brings us to the powerful combination of the existing VMware capabilities with recently integrated Lastline feature set:
NSX Network Detection and Response
Network Detection and Response (NDR) is a category of security solutions that complement EDR (we talk about Endpoint Detection and Response later) tools.
Powered by artificial intelligence (AI), NSX NDR maps and defends against MITRE ATT&CK techniques with the current capabilities:
NSX NDR protects the network, cloud and hybrid cloud traffic, and provides a cloud-based and on-prem architecture that enables sensors to gain comprehensive visibility into traffic that crosses the network perimeter (north/south), as well as traffic that moves laterally inside the perimeter (east/west).
NSX NDR uses a combination of four complementary technologies to detect and analyze advanced threats:
Behavior-based Network Traffic Analysis (NTA)
Network Traffic Analysis tools are all about detecting anomalies within the network (on-prem and public cloud) and use AI to create models of normal network activity and then alert on anomalies.
The challenge today is that not all anomalies are malicious. With Lastline’s NTA, VMware can now pick up threat behaviors and correlate these to network anomalies and vice versa. Because of this, according to VMware, they have the industry’s most accurate threat detection with minimal false positives.
Intrusion Detection and Prevention System (IDPS)
The NSX Advanced Threat Protection bundle includes IDS/IPS, which is integrated into NSX. The NSX Distributed IDS/IPS benefits from the unique application context from the hypervisor and network virtualization layers to make threat detection more accurate, efficient and dynamic.
The key capabilities of NSX Distributed IDS/IPS include:
Included with NSX Advanced Threat Prevention, Advanced Threat Analyzer provides complete malware analysis and enables accurate detection and prevention of advanced threats. It deconstructs every behavior engineered into a file or URL, and sees all instructions that a program executes, all memory content, and all operating system activity.
Other malware detection technologies, such as traditional sandboxes, only have visibility down to the operating system level. They can inspect content and identify potentially malicious code, but they can’t interact with malware like NSX Advanced Threat Analyzer can. As a result, they have significantly lower detection rates and higher false positives, in addition to being easily identified and evaded by advanced malware. (Advanced threats evade other sandboxing technologies by recognizing the sandbox environment or using kernel-level exploits.)
The VMware Threat Analysis Unit automatically shares the malware characteristics, behaviors and associated IoCs (Indicator of Compromises) of every malicious object curated and analyzed by VMware with all VMware customers and partners.
NSX Advanced Threat Analyzer continuously updates the VMware TAU in real time with intelligence from partner and customer environments around the world.
NSX Firewall for Baremetal Hosts. For organizations needing an agent-based network segmentation solution.
NSX Firewall. For organizations with one or more sites (optionally including public cloud endpoints) that primarily need advanced security services, select advanced networking capabilities, and traffic flow visibility and security operations with NSX Intelligence.
NSX Firewall with Advanced Threat Protection. For organizations that need NSX Firewall capabilities as well as advanced threat prevention capabilities, such as IDS/IPS, threat analysis, and network detection and response.
Use Case with Network Virtualization
If you are a customer with a NSX Data Center Advanced or Enterprise+ license, who uses NSX for network virtualization only today, you just need the “NSX ATP add-on” for NSX Data Center Advanced or Enterprise+.
Note: The ATP add-on requires NSX-T 3.1 and above.
Use Case without Network Virtualization (no NSX Data Center)
If you have no need for network virtualization for now, you have the following options:
If you look for base firewall features, you can get started with the NSX Firewall license.
Should you look for base firewall features plus advanced threat protection, then start with NSX Firewall with Advanced Threat Protection.
From here you still can down the network virtualization path and get the NSX Data Center Enterprise+ add-on for ATP
Use Case for VCF Customers
VCF customers have the option to start with the NSX ATP add-on for NSX NDC Adv/Ent+ as well.
Carbon Black Endpoint Detection and Response (EDR)
Before the Carbon Black acquisition, VMware already had strong technology, but was not seen or known as cybersecurity vendor. And it was really this acquisition that made the whole industry understand that VMware had to be taken seriously now as a security vendor.
So, what is EDR according to Wikipedia?
“Endpoint detection and response technology is used to protect endpoints, which are computer hardware devices, from threat. Creators of the EDR technology-based platforms deploy tools to gather data from endpoint devices, and then analyze the data to reveal potential cyber threats and issues. It is a protection against hacking attempts and theft of user data. The software is installed on the end-user device and it is continually monitored. The data is stored in a centralized database. In an incident when a threat is found, the end-user is immediately prompted with preventive list of actions.”
EDR is essential since local activities on machines that may be malicious are not visible on the network. VMware Carbon Black EDR is an incident response and threat hunting solution designed for security operations centers (SOCs) and incident response (IR) teams. Enterprise EDR is delivered through the VMware Carbon Black Cloud, an endpoint protection platform that consolidates security in the cloud using a single agent, console and dataset.
The Lastline acquisition, which came after Carbon Black, was just another brilliant move from VMware!
XDR – VMware Security brings together EDR and NDR
Again, while EDR protects endpoints, NDR protects the network, so that an organization’s entire IT infrastructure is secured. EDR gives security professionals visibility into endpoints that might be compromised, but this isn’t enough when an attack has moved across the network and into other systems by the time the security team is aware of it.
This is where XDR comes in. VMware rolled out its Extended Detection and Response (XDR) strategy at VMworld 2020. By the way, it was in 2020 when Gartner named XDR as one of the top nine cybersecurity trends.
By providing a holistic view of activity across the system that avoids visibility gaps, XDR allows security teams to understand where a threat comes from and how it’s spreading across the environment – in order to eliminate it. In other words, XDR offers greater analysis and correlation capabilities and a holistic point of view.
VMware’s XDR platform is the Carbon Black Cloud. Carbon Black Cloud’s evolution into an XDR platform includes product integrations with existing VMware products like Workspace ONE, vSphere and the NSX service-defined firewall, as well as third-party partner platforms.
At the Carbon Black Connect 2020 event, VMware announced launched their Next-Gen SOC Alliance that features integrations with the VMware Carbon Black Cloud to deliver key XDR capabilities and context into Security Information and Event Management (SIEM) technologies.
We’re in an epic war against cybercrime. We know the asymmetric nature of this war – you will not win by trying to staff your SOC with more analysts. Nor can the battle be won by deploying an individual technology focused on only one part of your IT infrastructure. EDR and NDR along with your SIEM form the winning combination you need to win the war.
The Carbon Black acquisition gave VMware a strong cybersecurity foundation to build on. The recent acquisition of Lastline VMware added sandboxing and network traffic analysis capabilities to their internal firewall, which is provided by NSX.
I don’t think it’s about “can VMware become a leading cybersecurity vendor” anymore. VMware has the most advanced internal firewall and is already becoming a leading cybersecurity vendor. The recent Global InfoSec award just confirms this statement:
Most Innovative in Endpoint Security” for VMware Carbon Black Cloud
“Market Leader in Firewall” for VMware NSX Service-defined Firewall
My last article focused on application modernization and data portability in a multi-cloud world. I explained the value of the VMware Tanzu portfolio by mentioning a consistent infrastructure and consistent application platform approach, which ultimately delivers a consistent developer experience. I also dedicated a short section about Tanzu Service Mesh, which is only one part of the unified Tanzu control plane (besides Tanzu Mission Control and Tanzu Observability) for multiple Kubernetes clusters and clouds.
When you hear or see someone writing about TSM, you very soon get to the point, where the so-called “Global Namespaces” (GNS) are being mentioned, which has the magic power to stitch hybrid applications together that run in multiple clouds.
Believe me when I say that Tanzu Service Mesh (TSM) is rising and becoming the next superstar of the VMware portfolio. I think Joe Baguley would agree here. 😀
Note: Sticking “Mesh” on things is the new cool. Yours, Joe Baguley-Mesh.
Before we start talking about Tanzu Service Mesh and the magical power of Global Namespaces, let us have a look at the term “Namespaces” first.
Namespaces give you a way to organize clusters into virtual carved out sub-clusters, which can be helpful when different teams, tenants or projects share the same Kubernetes cluster. This form of a namespace provides a method to better share resources, because it ensures fair allocation of these resources with the right permissions.
So, using namespaces gives you a way of isolation that developers never affect other project teams. Policies allow to configure compute resources by defining resource quotas for CPU or memory utilization. This also ensures the performance of a specific namespace, its resources (pods, services etc.) and the Kubernetes cluster in general.
Although namespaces are separate from each other, they can communicate with each other. Network policies can be configured to create isolated and non-isolated pods. For example, a network policy can allow or deny all traffic coming from other namespaces.
Ellei Mei explained this in a very easy in her article after Project Pacific had been made public in September 2019:
Think of a farmer who divides their field (cluster + cluster resources) into fenced-off smaller fields (namespaces) for different herds of animals. The cows in one fenced field, horses in another, sheep in another, etc. The farmer would be like operations defining these namespaces, and the animals would be like developer teams, allowed to do whatever they do within the boundaries they are allocated.
The first time I heard of Kubernetes or vSphere Namespaces was in fact at VMworld 2019 in Barcelona. VMware then presented a new app-focused management concept. This concept described a way to model modern application and all their parts, and we call this a vSphere Namespace today.
With Project Pacific (today known vSphere with Tanzu or Tanzu Kubernetes Grid), VMware went one step further and extended the Kubernetes Namespace by adding more options for compute resource allocation, vMotion, encryption, high availability, backup & restore, and snapshots.
Rather than having to deal with each namespace and its containers, vSphere Namespaces (also called “guardrails” sometimes) can draw a line around the whole application and services including virtual machines.
With the re-architecture of vSphere and the integration of Kubernetes as its control plane, namespaces can be seen as the new unit of management.
Imagine that you might have thousands of VMs in your vCenter inventory that you needed to deal with. After you group those VMs into their logical applications, you may only have to deal with dozens of namespaces now.
If you need to turn on encryption for an application, you can just click a button on the namespace in vCenter and it does it for you. You don’t need to deal with individual VMs anymore.
vSphere Virtual Machine Service
With the vSphere 7 Update 2a release, VMware provided the “VM Service” that enables Kubernetes-native provisioning and management of VMs.
For many organizations legacy applications are not becoming modern over night, they become hybrid first before the are completely modernized. This means we have a combination of containers and virtual machines forming the application, and not find containers only. I also call this a hybrid application architecture in front of my customers. For example, you may have a containerized application that uses a database hosted in a separate VM.
So, developers can use the existing Kubernetes API and a declarative approach to create VMs. No need to open a ticket anymore to request a virtual machine. We talk self-service here.
Tanzu Mission Control – Namespace Management
Tanzu Mission Control (TMC) is a VMware Cloud (SaaS) service that provides a single control point for multiple teams to remove the complexities from managing Kubernetes cluster across multiple clouds.
One of the ways to organize and view your Kubernetes resources with TMC is by the creation of “Workspaces”.
Workspaces allows you to organize your namespaces into logical groups across clusters, which helps to simplify management by applying policies at a group level. For example, you could apply an access policy to an entire group of clusters (from multiple clouds) rather than creating separate policies for each individual cluster.
Think about backup and restore for a moment. TMC and the concept of workspaces allow you to back up and restore data resources in your Kubernetes clusters on a namespace level.
Management and operations with a new application view!
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 have to think about the connectivity and communication between your app’s microservices.
One of the main ideas and features behind a service mesh was to provide service-to-service communication for distributed applications running in multiple Kubernetes clusters hosted in different private or public clouds.
The number of Kubernetes service meshes has rapidly increased over the last few years and has gotten a lot of hype. No wonder why there are different service mesh offerings around:
AWS Apps Mesh
OpenShift Service Mesh by Red Hat
Open Service Mesh AKS add-on (currently preview on Azure)
Istio is probably the most famous one on this list. For me, it is definitely the one my customers look and talk about the most.
Service mesh brings a new level of connectivity between services. With service mesh, we inject a proxy in front of each service; in Istio, for example, this is done using a “sidecar” within the pod.
Istio’s architecture is divided into a data plane based on Envoy (the sidecar) and a control plane, that manages the proxies. With Istio, you inject the proxies into all the Kubernetes pods in the mesh.
As you can see on the image, the proxy sits in front of each microservice and all communications are passed through it. When a proxy talks to another proxy, then we talk about a service mesh. Proxies also handle traffic management, errors and failures (retries) and collect metric for observability purposes.
Challenges with Service Mesh
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.
Istio supports a so-called multi-cluster deployment with one service mesh stretched across Kubernetes clusters, but you’ll end up with a stretched Istio control plane, which eliminates the independence of each cluster.
So, a lot of customers also talk about better and easier manageability without dependencies between clouds and different Kubernetes clusters from different vendors.
That’s the moment when Tanzu Service Mesh becomes very interesting. 🙂
Tanzu Service Mesh (formerly known as NSX Service Mesh)
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.
When onboarding a new cluster on Tanzu Service Mesh, the service deploys a curated version of Istio signed and supported by VMware. This Istio deployment is the same as the upstream Istio in every way, but it also includes an agent that communicates with the Tanzu Service Mesh global control plane. Istio installation is not the most intuitive, but the onboarding process of Tanzu Service Mesh simplifies the process significantly.
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.
Global Namespaces (GNS)
Yep, another kind of a namespace, but the most exciting one! 🙂
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.
The following diagram represents the global namespace concept and other pieces in a high-level architectural view. The components of one application are distributed in two different Kubernetes clusters: one of them is on-premises and the other in a public cloud. The Global Namespace creates a logical view of these application components and provides a set of basic services for the components.
If we take application continuity as another example for a use case, we would deploy an app in more than one cluster and possibly in a remote region for disaster recovery (DR), with a load balancer between the locations to direct traffic to both clusters. This would be an active-active scenario. With Tanzu Service Mesh, you could group the clusters into a Global Namespace and program it to automatically redirect traffic in case of a failure.
In addition to the use case and support for multi-zone andmulti-region high availability and disaster recovery, you can also provide resiliency with automated scaling based on defined Service-Level Objectives (SLO) for multi-cloud apps.
VMware Modern Apps Connectivity Solution
In May 2021 VMware introduced a new solution that brings together the capabilities of Tanzu Service Mesh and NSX Advanced Load Balancer (NSX ALB, formerly Avi Networks) – not only for containers but also for VMs. While Istio’s Envoy only operates on layer 7, VMware provides layer 4 to layer 7 services with NSX (part of TSM) and NSX ALB, which includes L4 load balancing, ingress controllers, GSLB, WAF and end-to-end service visibility.
This solution speeds the path to app modernization with connectivity and better security across hybrid environments and hybrid app architectures.
One thing I can say for sure: The future for Tanzu Service Mesh is bright!
Many customers are looking for ways for offloading security (encryption, authentication, authorization) from an application to a service mesh.
One great example and use case from the financial services industry is crypto agility, where a “crypto service mesh” (a specialized service mesh) could be part of a new architecture, which provides quantum-safe certificates.
And when we offload encryption, calculation, authentication etc., then we may have other use cases for SmartNICs and Project Monterey.
It was 2019 when VMware announced Tanzu and Project Pacific. A lot has happened since then and almost everyone is talking about application modernization nowadays. With my strong IT infrastructure background, I had to learn a lot of new things to survive initial conversations with application owners, developers and software architects. And in the same time VMware’s Kubernetes offering grew and became very complex – not only for customers, but for everyone I believe. 🙂
I already wrote about VMware’s vision with Tanzu: To put a consistent “Kubernetes grid” over any cloud
This is the simple message and value hidden behind the much larger topics when discussing application modernization and application/data portability across clouds.
The goal of this article is to give you a better understanding about the real value of VMware Tanzu and to explain that it’s less about Kubernetes and the Kubernetes integration with vSphere.
Before we can talk about the modernization of applications or the different migration approaches like:
Retain – Optimize and retain existing apps, as-is
Rehost/Migration (lift & shift) – Move an application to the public cloud without making any changes
Replatform (lift and reshape) – Put apps in containers and run in Kubernetes. Move apps to the public cloud
Rebuild and Refactor – Rewrite apps using cloud native technologies
Retire – Retire traditional apps and convert to new SaaS apps
…we need to have a look at the palette of our applications:
Big Data – Splunk, Elasticsearch, ELK stack, Greenplum, Kafka, Hadoop
In an app modernization discussion, we very quickly start to classify applications as microservices or monoliths. From an infrastructure point of view you look at apps differently and call them “stateless” (web apps) or “stateful” (SQL, NoSQL, Big Data) apps.
And with Kubernetes we are trying to overcome the challenges, which come with the stateful applications related to app modernization:
What does modernization really mean?
How do I define “modernization”?
What is the benefit by modernizing applications?
What are the tools? What are my options?
What has changed? Why is everyone talking about modernization? Why are we talking so much about Kubernetes and cloud native? Why now?
To understand the benefits (and challenges) of app modernization, we can start looking at the definition from IBM for a “modern app”:
“Application modernization is the process of taking existing legacy applications and modernizing their platform infrastructure, internal architecture, and/or features. Much of the discussion around application modernization today is focused on monolithic, on-premises applications—typically updated and maintained using waterfall development processes—and how those applications can be brought into cloud architecture and release patterns, namely microservices“
Modern applications are collections of microservices, which are light, fault tolerant and small. Microservices can run in containers deployed on a private or public cloud.
Which means, that a modern application is something that can adapt to any environment and perform equally well.
Note: App modernization can also mean, that you must move your application from .NET Framework to .NET Core.
I have a customer, that is just getting started with the app modernization topic and has hundreds of Windows applications based on the .NET Framework. Porting an existing .NET app to .NET Core requires some work, but is the general recommendation for the future. This would also give you the option to run your .NET Core apps on Windows, Linux and macOS (and not only on Windows).
A modern application is something than can run on bare-metal, VMs, public cloud and containers, and that easily integrates with any component of your infrastructure. It must be something, that is elastic. Something, that can grow and shrink depending on the load and usage. Since it is something that needs to be able to adapt, it must be agile and therefore portable.
Cloud Native Architectures and Modern Designs
If I ask my VMware colleagues from our so-called MAPBU (Modern Application Platform Business Unit) how customers can achieve application portability, the answer is always: “Cloud Native!”
Many organizations and people see cloud native as going to Kubernetes. But cloud native is so much more than the provisioning and orchestration of containers with Kubernetes. It’s a about collaboration, DevOps, internal processes and supply chains, observability/self-healing, continuous delivery/deployment and cloud infrastructure.
There are so many definitions around “cloud native”, that Kamal Arora from Amazon Web Services and others wrote the book “Cloud Native Architecture“, which describes a maturity model. This model helps you to understand, that cloud native is more a journey than only restrictive definition.
The adoption of cloud services and applying an application-centric design are very important, but the book also mentions that security and scalability rely on automation. And this for example could bring the requirement for Infrastructure as Code (IaC).
In the past, virtualization – moving from bare-metal to vSphere – didn’t force organizations to modernize their applications. The application didn’t need to change and VMware abstracted and emulated the bare-metal server. So, the transition (P2V) of an application was very smooth and not complicated.
And this is what has changed today. We have new architectures, new technologies and new clouds running with different technology stacks. We have Kubernetes as framework, which requires applications to be redesigned for these platforms.
That is the reason why enterprises have to modernize their applications.
One of the “five R’s” mentioned above is the lift and shift approach. If you don’t want or need to modernize some of your applications, but move to the public cloud in an easy, fast and cost efficient way, have a look at VMware’ hybrid cloud extension (HCX).
In this article I focus more on the replatform and refactor approaches in a multi-cloud world.
Kubernetize and productize your applications
Assuming that you also define Kubernetes as the standard to orchestrate your containers where your microservices are running in, usually the next decision would be about the Kubernetes “product” (on-prem, OpenShift, public cloud).
Talking to my customers, most of them mention the storage and network integration as one of their big challenges with Kubernetes. Their concern is about performance, resiliency, different storage and network patterns, automation, data protection/replication, scalability and cloud portability.
Why do organizations need portability?
There are many use cases and requirements that portability (infrastructure independence) becomes relevant. Maybe it’s about a hardware refresh or data center evacuation, to avoid vendor/cloud lock-in, not enough performance with the current infrastructure or it could be about dev/test environments, where resources are deployed and consumed on-demand.
Multi-Cloud Application Portability with VMware Tanzu
To explore the value of Tanzu, I would like to start by setting the scene with the following customer use case:
On-premises: VMware vSphere infrastructure, no containerization yet, only legacy applications
In this case the customer is following a cloud-appropriate approach to define which cloud is the right landing zone for their applications. They decided to develop new applications in the public cloud and use the native services from Azure and AWS. The customers still has hundreds of legacy applications (monoliths) on-premises and didn’t decide yet, if they want to follow a “lift and shift and then modernize” approach to migrate a number applications to the public cloud.
But some of their application owners already gave the feedback, that their applications are not allowed to be hosted in the public cloud, have to stay on-premises and need to be modernized locally.
At the same time the IT architecture team receives the feedback from other application owners, that the journey to the public cloud is great on paper, but brings huge operational challenges with it. So, IT operations asks the architecture team if they can do something about that problem.
Both cloud operations for Azure and AWS teams deliver a different quality of their services, changes and deployments take longer with one of their public clouds, they have problems with overlapping networks, different storage performance characteristics and APIs.
Another challenge is the role-based access to the different clouds, Kubernetes clusters and APIs. There is no central log aggregation and no observability (intelligent monitoring & alerting). Traffic distribution and load balancing are also other items on this list.
Because of the feedback from operations to architecture, IT engineering received the task to define a multi-cloud strategy, that solves this operational complexity.
Notes: These are the regular multi-cloud challenges, where clouds are the new silos and enterprises have different teams with different expertise using different management and security tools.
This is the time when VMware’s multi-cloud approach Tanzu become very interesting for such customers.
Consistent Infrastructure and Management
The first discussion point here would be the infrastructure. It’s important, that the different private and public clouds are not handled and seen as silos. VMware’s approach is to connect all the clouds with the same underlying technology stack based on VMware Cloud Foundation.
Beside the fact, that lift and shift migrations would be very easy now, this approach brings two very important advantages for the containerized workloads and the cloud infrastructure in general. It solves the challenge with the huge storage and networking ecosystem available for Kubernetes workloads by using vSAN and NSX Data Center in any of the existing clouds. Storage and networking and security are now integrated and consistent.
For existing workloads running natively in public clouds, customers can use NSX Cloud, which uses the same management plane and control plane as NSX Data Center. That’s another major step forward.
Consistent Application Platform and Developer Experience
Looking at organization’s application and container platforms, achieving consistent infrastructure is not required, but obviously very helpful in terms of operational and cost efficiency.
To provide a consistent developer experience and to abstract the underlying application or Kubernetes platform, you would follow the same VMware approach as always: to put a layer on top.
Here the solution is called Tanzu Kubernetes Grid (TKG), that provides a consistent, upstream-compatible implementation of Kubernetes, that is tested, signed and supported by VMware.
A Tanzu Kubernetes cluster is an opinionated installation of Kubernetes open-source software that is built and supported by VMware. In all the offerings, you provision and use Tanzu Kubernetes clusters in a declarative manner that is familiar to Kubernetes operators and developers. The different Tanzu Kubernetes Grid offerings provision and manage Tanzu Kubernetes clusters on different platforms, in ways that are designed to be as similar as possible, but that are subtly different.
VMware Tanzu Kubernetes Grid (TKG aka TKGm)
Tanzu Kubernetes Grid can be deployed across software-defined datacenters (SDDC) and public cloud environments, including vSphere, Microsoft Azure, and Amazon EC2. I would assume, that the Google Cloud is a roadmap item.
TKG allows you to run Kubernetes with consistency and makes it available to your developers as a utility, just like the electricity grid. TKG provides the services such as networking, authentication, ingress control, and logging that a production Kubernetes environment requires.
This TKG version is also known as TKGm for “TKG multi-cloud”.
VMware Tanzu Kubernetes Grid Service (TKGS aka vSphere with Tanzu)
TKGS is the option vSphere admins want to hear about first, because it allows you to turn a vSphere cluster to a platform running Kubernetes workloads in dedicated resources pools. TKGS is the thing that was known as “Project Pacific” in the past.
Once enabled on a vSphere cluster, vSphere with Tanzu creates a Kubernetes control plane directly in the hypervisor layer. You can then run Kubernetes containers by deploying vSphere Pods, or you can create upstream Kubernetes clusters through the VMware Tanzu Kubernetes Grid Service and run your applications inside these clusters.
VMware Tanzu Mission Control (TMC)
In our use case before, we have AKS and EKS for running Kubernetes clusters in the public cloud.
The VMware solution for multi-cluster Kubernetes management across clouds is called Tanzu Mission Control, which is a centralized management platform for the consistency and security the IT engineering team was looking for.
Available through VMware Cloud Services as SaaS offering, TMC provides IT operators with a single control point to provide their developers self-service access to Kubernetes clusters.
TMC also provides cluster lifecycle management for TKG clusters across environment such as vSphere, AWS and Azure.
It allows you to bring the clusters you already have in the public clouds or other environments (with Rancher or OpenShift for example) under one roof via the attachment of conformant Kubernetes clusters.
Not only do you gain global visibility across clusters, teams and clouds, but you also get centralized authentication and authorization, consistent policy management and data protection functionalities.
VMware Tanzu Observability by Wavefront (TO)
Tanzu Observability extends the basic observability provided by TMC with enterprise-grade observability and analytics.
Wavefront by VMware helps Tanzu operators, DevOps teams, and developers get metrics-driven insights into the real-time performance of their custom code, Tanzu platform and its underlying components. Wavefront proactively detects and alerts on production issues and improves agility in code releases.
TO is also a SaaS-based platform, that can handle the high-scale requirements of cloud native applications.
VMware Tanzu Service Mesh (TSM)
Tanzu Service Mesh, formerly known as NSX Service Mesh, provides consistent connectivity and security for microservices across all clouds and Kubernetes clusters. TSM can be installed in TKG clusters and third-party Kubernetes-conformant clusters.
Organizations that are using or looking at the popular Calico cloud native networking option for their Kubernetes ecosystem often consider an integration with Istio (Service Mesh) to connect services and to secure the communication between these services.
The combination of Calico and Istio can be replaced by TSM, which is built on VMware NSX for networking and that uses an Istio data plane abstraction. This version of Istio is signed and supported by VMware and is the same as the upstream version. TSM brings enterprise-grade support for Istio and a simplified installation process.
One of the primary constructs of Tanzu Service Mesh is the concept of a Global Namespace (GNS). GNS allows developers using Tanzu Service Mesh, regardless of where they are, to connect application services without having to specify (or even know) any underlying infrastructure details, as all of that is done automatically. With the power of this abstraction, your application microservices can “live” anywhere, in any cloud, allowing you to make placement decisions based on application and organizational requirements—not infrastructure constraints.
Note: On the 18th of March 2021 VMware announced the acquisition of Mesh7 and the integration of Mesh7’s contextual API behavior security solution with Tanzu Service Mesh to simplify DevSecOps.
The VMware Tanzu portfolio comes with three different editions: Basic, Standard, Advanced
Tanzu Basic enables the straightforward implementation of Kubernetes in vSphere so that vSphere admins can leverage familiar tools used for managing VMs when managing clusters = TKGS
Tanzu Standard provides multi-cloud support, enabling Kubernetes deployment across on-premises, public cloud, and edge environments. In addition, Tanzu Standard includes a centralized multi-cluster SaaS control plane for a more consistent and efficient operation of clusters across environments = TKGS + TKGm + TMC
Tanzu Advanced builds on Tanzu Standard to simplify and secure the container lifecycle, enabling teams to accelerate the delivery of modern apps at scale across clouds. It adds a comprehensive global control plane with observability and service mesh, consolidated Kubernetes ingress services, data services, container catalog, and automated container builds = TKG (TKGS & TKGm) + TMC + TO + TSM + MUCH MORE
Tanzu Data Services
Another topic to reduce dependencies and avoid vendor lock-in would be Tanzu Data Services – a separate part of the Tanzu portfolio with on-demand caching (Tanzu Gemfire), messaging (Tanzu RabbitMQ) and database software (Tanzu SQL & Tanzu Greenplum) products.
Bringing all together
As always, I’m trying to summarize and simplify things where needed and I hope it helped you to better understand the value and capabilities of VMware Tanzu.
There are so many more products available in the Tanzu portfolio, that help you to build, run, manage, connect and protect your applications.
IT organizations are looking for consistent operations, which is enabled by consistent infrastructure. Public cloud providers like AWS and Microsoft offer an extension of their cloud infrastructure and native services to the private cloud and edge, which is also known as Data Center as a Service.
Amazon Web Services (AWS) provides a fully managed service with AWS Outposts, that offers AWS infrastructure, AWS services, APIs and their tools to any data center or on-premises facility.
Microsoft has Azure Stack is even working on a new Azure Stack hybrid cloud solution that is codenamed “Fiji” to provide the ability to run Azure as a managed local cloud.
What do these offerings have in common or why would customers choose one (or even both) of these hybrid cloud options?
They bring the public cloud operation model to the private cloud or edge in form of one or more racks and servers provided as a fully managed service.
AWS Outposts (generally available since December 2019) and Azure Stack Fiji (in development) provide the following:
Extension of the public cloud services to the private cloud and edge
Consistent infrastructure with consistent operations
Local processing of data (e.g., analytics at the data source)
Local data residency (governance and security)
Low latency access to on-premises systems
Local migrations and modernization of applications with local system interdependencies
Build, run and manage on-premises applications using existing and familiar services and tools
Modernize applications on-prem resp. at the edge
Prescriptive infrastructure and vendor managed lifecycle and maintenance (racks and servers)
Creation of different physical pools and clusters depending on your compute and storage needs (different form factors)
Same licensing and pricing options on-premises (like in the public cloud)
The pretty new AWS Outposts or the future Azure Stack Fiji solution are also called “Local Cloud as a Service” (LCaaS) or “Data Center as a Service” and meant to be consumed and delivered in the on-prem data center or at the edge. It’s about bringing the public cloud to your data center or edge location.
The next phase of cloud transformations is about the “edge” of an enterprise cloud and we know today that private and hybrid cloud strategies are critical for the implementation of IT infrastructure and the operation of it.
If you come from VMware’s standpoint, then it’s not about extending the public cloud to the local data centers. It’s about extending your VMware-based private cloud to the edge or the public cloud.
This article focuses on the local (private) cloud as a service options from VMware, not the public cloud offerings.
Before I describe the different VMware LCaaS offerings based on VMware Cloud Foundation, let me show and explain the different features and technologies my customers ask about when they plan to build a private cloud with public cloud characteristics in mind.
I work with customers from different verticals like
fast-moving consumer goods
which are hosting IT infrastructure in multiple data centers all over the world including hundreds of smaller locations. My customers belong to different vertical markets, but are looking for the same features and technologies when it comes to edge computing and delivering a managed cloud on-premises.
Compute and Storage. They are looking for pre-validated and standardized configuration offerings to meet their (application) needs. Most of them describe hardware blueprints with t-shirts sizes (small, medium, large). These different servers or instances provide different options and attributes, which should provide enough CPU, RAM, storage and networking capacity based on their needs. Usually you’ll find terms like “general purpose”, “compute optimized” or “memory optimized” node types or instances.
Networking. Most of my customers look for the possibility to extend their current network (aka elastic or cloud-scale networking) to any other cloud. They prefer a way to use the existing network and security policies and to provide software-defined networking (SDN) services like routing, firewalling and IDS/IPS, load balancing – also known as virtualized network functions (VNF). Service providers are also looking at network function virtualization (NFV), which includes emerging technologies like 5G and IoT. As cloud native or containerized applications become more important, service providers also discuss containerized network functions (CNF).
Services. Applications consist of one or many (micro-)services. All my conversations are application-centric and focus on the different application components. Most of my discussions are about containers, databases and video/data analytics at the edge.
Security. Customers, that are running workloads in the public cloud, are familiar with the shared responsibility model. The difference between public cloud and local cloud as a service offering is the physical security (racks, servers, network transits, data center access etc.).
Scalability and Elasticity. IT providers want to provide the simplicity and agility on-prem as their customers (the business) would expect it from a public cloud provider. Scalability is about a planned level of capacity that can grow or shrink as needed.
Resource Pooling and Sharing. Larger enterprises and service providers are interested in creating dedicated workload domains and resource clusters, but also look for a way to provide infrastructure multi-tenancy.
The challenge for today’s IT teams is, that edge locations are not often well defined. And these IT teams need an efficient way to manage different infrastructure sizes (can range from 2 nodes up to 16 or 24 nodes), for sometimes up to 400 edge locations.
Rethinking Private Clouds
Organizations have two choices when it comes to the deployment of a private cloud extension to the edge. They could continue using the current approach, which includes the design, deployment and operation of their own private cloud. Another pretty new option would be the subscription of a predefined “Data Center as a Service” offering.
Enterprises need to develop and implement a cloud strategy to support the existing workloads, which are still mostly running on VMware vSphere, and build something, which is vendor and cloud-agnostic. Something, that provides a (public) cloud exit strategy at the same time.
If you decide to go for AWS Outposts or the coming Azure Stack Fiji solution, which for sure are great options, how would you migrate or evacuate workloads to another cloud and technology stack?
VMware Cloud on Dell EMC
At VMworld 2019 VMware announced the general availability of VMware Cloud on Dell EMC (VMC on Dell EMC). In 2018 introduced as “Project Dimension”, the idea behind this concept was to deliver a (public) cloud experience to customers on-premises. Give customers the best of two worlds:
The simplicity, flexibility and cost model of the public cloud with the security and control of your private cloud infrastructure.
Initially, Project Dimension was focused primarily on edge use cases and was not optimized for larger data centers.
Note: This has changed with the introduction of the 2nd generation of VMC on Dell EMC in May 2020 to support different density and performance use cases.
VMC on Dell EMC is a VMware-managed service offering with these components:
A software-defined data center based von VMware Cloud Foundation (VCF) running on Dell EMC VxRail
ESXi, vSAN, NSX, vCenter Server
Dell servers, management & ToR switches, racks, UPS
Standby VxRail node for expansion (unlicensed)
Option for half or full-height rack
Multiple cluster support in a single rack
Clusters start with a minimum of 3 nodes (not 4 as you would expect from a regular VCF deployment)
VMware SD-WAN (formerly known as VeloCloud) appliances for remote management purposes only at the moment
Customer self-service provisioning through cloud.vmware.com
Maintenance, patching and upgrades of the SDDC performed by VMware
Maintenance, patching and upgrades of the Dell hardware performed by VMware (Dell provides firmware, drivers and BIOS updates)
1- or 3-year term subscription commitment (like with VMC on AWS)
There is no “one size fits all” when it comes to hosting workloads at the edge and in your data centers. VMC on Dell EMC provides also different hardware node types, which should match with your defined t-shirt sizes (blueprints).
If we talk about at a small edge location with a maximum of 5 server nodes, you would go for a half-height rack. The full-height rack can host up to 24 nodes (8 clusters). Currently, the largest instance type would be a good match for high density, storage hungry workloads such as VDI deployments, databases or video analytics.
As HCX is part of the offering, you have the right tool and license included to migrate workloads between vSphere-based private and public clouds.
The following is a list of some VMworld 2020 breakout sessions presented by subject matter experts and focused on VMware Cloud on Dell EMC:
VMware Cloud Foundation and HPE Synergy with HPE GreenLake
At VMworld 2019 VMware announced that VMware Cloud Foundation will be offered in HPE’s GreenLake program running on HPE Synergy composable infrastructure (Hybrid Cloud as a Service). This gives VMware customers the opportunity to build a fully managed private cloud with the public cloud benefits in an on-premises environment.
HPE’s vision is built on a single platform that can span across multiple clouds and GreenLake brings the cloud consumption model to joint HPE and VMware customers.
If you are an AWS customer and look for a consistent hybrid cloud experience, then you would consider AWS Outposts.
There is also VMware variant of AWS Outposts available for customers, who already run their on-premises workloads on VMware vSphere or in a cloud vSphere-based environment running on top of the AWS global infrastructure (called VMC on AWS).
VMware Cloud on AWS Outposts is a on-premises as-a-service offering based on VMware Cloud Foundation. It integrates VMware’s software-defined data center software, including vSphere, vSAN and NSX. Ths Cloud Foundation stack runs on dedicated elastic Amazon EC2 bare-metal infrastructure, delivered on-premises with optimized access to local and remote AWS services.
Key capabilities and use cases:
Use familiar VMware tools and skillsets
No need to rewrite applications while migrating workloads
Direct access to local and native AWS services
Service is sold, operated and supported by VMware
VMware as the single point of primary contact for support needs, supplemented by AWS for hardware shipping, installation and configuration
Host-level HA with automated failover to VMware Cloud on AWS
Resilient applications required to work in the event of WAN link downtime
Application modernization with access to local and native AWS services
1- or 3-year term subscription commitment
42U AWS Outposts rack, fully assembled and installed by AWS (including ToR switches)
Minimum cluster size of 3 nodes (plus 1 dark node)
Current cluster maximum of 16 nodes
Currently, VMware is running a VMware Cloud on AWS Outposts Beta program, that lets you try the pre-release software on AWS Outposts infrastructure. An early access program should start in the first half of 2021, which can be considered as a customer paid proof of concept intended for new workloads only (no migrations).
VMware on Azure Stack
To date there are no plans communicated by Microsoft or VMware to make Azure VMware Solution, the vSphere-based cloud offering running on top of Azure, available on-premises on the current or future Azure Stack family.
VMware on Google Anthos
To date there are no plans communicated by Google or VMware to make Google Cloud VMware Engine, the vSphere-based cloud offering running on top of the Google Cloud Platform (GCP), available on-premises.
The only known supported combination of a Google Cloud offering running VMware on-premises is Google Anthos (Google Kubernetes Engine on-prem).
Multi-Cloud Application Portability
Multi-cloud is now the dominant cloud strategy and many of my customers are maintaining a vSphere-based cloud on-premises and use at least two of the big three public clouds (AWS, Azure, Google).
Following a cloud-appropriate approach, customers are inspecting each application and decide which cloud (private or public) would be the best to run this application on. VMware gives customers the option to run the Cloud Foundation technology stack in any cloud, which doesn’t mean, that customers at the same time are not going cloud-native and still add AWS and Azure to the mix.
How can I achieve application portability in a multi-cloud environment when the underlying platform and technology formats differ from each other?
This is a question I hear a lot. Kubernetes is seen as THE container orchestration tool, which at the same time can abstract multiple public clouds and the complexity that comes with them.
A lot of people also believe that Kubernetes is enough to provide application portability and figure out later, that they have to use different Kubernetes APIs and management consoles for every cloud and Kubernetes (e.g., Rancher, Azure, AWS, Google, RedHat OpenShift etc.) flavor they work with.
That’s the moment we have to talk about VMware Tanzu and how it can simplify things for you.
The Tanzu portfolio provides the next generation the building blocks and steps for modernizing your existing workloads while providing capabilities of Kubernetes. Additionally, Tanzu also has broad support for containerization across the entire application lifecycle.
Tanzu gives you the possibility to build, run, manage, connect and protect applications and to achieve multi-cloud application portability with a consistent platform over any cloud – the so-called “Kubernetes grid”.
Note: I’m not talking about the product “Tanzu Kubernetes Grid” here!
I’m talking about the philosophy to put a virtual application service layer over your multi-cloud architecture, which provides a consistent application platform.
Tanzu Mission Control is a product under the Tanzu umbrella that provides central management and governance of containers and clusters across data centers, public clouds, and edge.
Enterprises must be able to extend the value of their cloud investments to the edge of the organization.
The edge is just one piece of a bigger picture and customers are looking for a hybrid cloud approach in a multi-cloud world.
Solutions like VMware Cloud on Dell EMC or running VCF on HPE Synergy with HPE Greenlake are only the first steps towards innovation in the private cloud and to bring the cost and operation model from the public cloud to the enterprises on-premises.
IT organizations are rather looking for ways to consume services in the future and care less about building the infrastructure or services by themselves.
The two most important differentiators for selecting an as-a-service infrastructure solution provider will be the provider’s ability to enable easy/consistent connectivity and the provider’s established software partner portfolio.
In cases where IT organizations want to host a self-managed data center or local cloud, you can expect, that VMware is going to provide a new and appropriate licensing model for it.
One of the reasons and one of the essential characteristics of a cloud computing model I mentioned is resource pooling.
By the National Institute of Standards and Technology (NIST) resource pooling is defined with the following words:
The provider’s computing resources are pooled to serve multiple consumers using a multi-tenant model, with different physical and virtual resources dynamically assigned and reassigned according to consumer demand. There is a sense of location independence in that the customer generally has no control or knowledge over the exact location of the provided resources but may be able to specify location at a higher level of abstraction (e.g., country, state, or data center).
This time I would like to focus on multi-tenancy and how you can achieve that on top of VMware Cloud Foundation (VCF) with Cloud Director (formerly known as vCloud Director) and vRealize Automation, which both could be part of a VMware cloud management platform (CMP).
There are many understandings around about multi-tenancy and different people have different definitions for it.
If we start from the top of an IT infrastructure, we will have application or software multi-tenancy with a single instance of an application serving multiple tenants. And in the past even running on the same virtual or physical server. In this case the multi-tenancy feature is built into the software, which is commonly accessed by a group of users with specific permissions. Each tenant gets a dedicated or isolated share of this application instance.
Coming from the bottom of the data center, multi-tenancy describes the isolation of resources (compute, storage) and networks to deliver applications. The best example here are (cloud) services providers.
Their goal is to create and provide virtual data centers (VDC) or a virtual private cloud (VPC) on top of the same physical data center infrastructure – for different tenants aka customers. Normally, the right VMware solution for this requirement and service providers would be Cloud Director, but this is maybe not completely true anymore with the release of vRealize Automation 8.x.
To make it easier for all of us, I’ll call Cloud Director and vCloud Director “vCD” from now on.
VMware Cloud Director (formerly vCloud Director)
Cloud Director is a product exclusively for cloud service providers via the VMware Cloud Provider Program (VCPP). Originally released in 2010, it enables service providers (SPs) to provision SDDC (Software-Defined Data Center) services as complete virtual data centers. vCD also keeps resources from different tenants isolated from each other.
Within vCD a unit of tenancy is called Organization VDC (OrgVDC). It is defined as a set of dedicated compute (CPU, RAM), storage and network resources. A tenant can be bound to a single OrgVDC or can be composed of multiple Organization VDCs. This is typically known as Infrastructure as a Service (IaaS).
A provider virtual data center (PVDC) is a grouping of compute, storage, and network resources from a single vCenter Server instance. Multiple organizations/tenants can share provider virtual data center resources.
Important: I assume that you are familiar with VCF, its core components (ESXi, vSAN, NSX, SDDC Manager) and architecture models (standard as the preferred).
Cloud Director components are currently not part of the VCF lifecycle automation, but it is a roadmap item!
Cloud Director Resource Hosting Models
vCD offers multiple hosting models:
In the shared hosting model, multiple tenant workloads run all together on the same resource groups without any performance assurance
In the reserved hosting model, performance of workloads is assured by resource reservation.
In the physical hosting model, hardware is dedicated to a single tenant and performance is assured by the allocated hardware
Tenant Using Shared Hosting on VCF Workload Domain
In this use case a tenant is using shared hosting backed by a VMware Cloud Foundation workload domain. A workload domain, which is mapped to a provider VDC.
Tenant Using Shared Hosting and Reserved Hosting on Multiple VCF Workload Domains
This use case describes the example of customer using shared and reserved hosting backed by multiple VCD workload domains. Here each cluster has a single resource pool mapped to a single PVDC.
Tenant Using Physical Hosting and Central Point of Management (CPOM)
The last example shows a single customer using physical hosting. You will notice that there is also a vSphere with Kubernetes workload domain. VMware Cloud Foundation automates the installation of vSphere with Kubernetes (Tanzu) which makes it incredibly easy to deploy and manage.
You can see that there is an “SDDC” box on top of the Kubernetes Cluster vCenter, which is attached to the “SDDC Proxy” entity. vCD can act as an HTTP/S proxy server between tenants and the underlying vSphere environment in VMware Cloud Foundation. An SDDC proxy is an access point to a component from an SDDC, for example, a vCenter Server instance, an ESXi host, or an NSX Manager instance.
The vCD becomes the central point of management (CPOM) in this case and the customer gets a complete dedicated SDDC with vCenter access.
Note: Since vCD 9.7 it is possible to present for example a vCenter Server instance securely to a tenant’s organization using the Cloud Director user interface. This is how you could build your own VMC-on-AWS-like cloud offering!
All 3 Tenants Together
Finally, we put it all together. In the first use case we can see that different customers are sharing resources from a single PVDC. We can also see that resources from a single vCenter can be split across different provider virtual datacenters and that we can mix and match multi-tenants workload domains and workload domains offering dedicated private cloud all together.
Cloud Director Service and VMware Cloud on AWS
If you don’t want to extend or operate your own data center or cloud infrastructure anymore and provide a managed service to multiple customer, there are still options for you available backed by VMware Cloud Foundation as well.
VMware sees not only new, but also existing VCPP partners moving towards a mixed-asset portfolio, where their cloud management platform consists of a VCPP and MSP (VMware SaaS offerings) contract. This allows them for example to run vCD on-premises for their current customers and the onboarding of new tenants would happen in the public cloud with CDS and VMC on AWS.
Enterprise Multi-Tenancy with vRealize Automation
With the release of vRealize Automation 8.1 (vRA) VMware offered support for dedicated infrastructure multi-tenancy, created and managed through vRealize Suite Lifecycle Manager. This means vRealize Automation enables customers or IT providers to set up multiple tenants or organizations within each deployment.
Providers can set up multiple tenant organizations and allocate infrastructure. Each tenant manages its own projects (team structures), resources and deployments.
Enabling tenancy creates a new Provider (default) organization. The Provider Admin will create new tenants, add tenant admins, setup directory synchronization, and add users. Tenant admins can also control directory synchronization for their tenant and will grant users access to services within their tenant. Additionally, tenant admins will configure Policies, Governance, Cloud Zones, Profiles, access to content and provisioned resources; within their tenant. A single shared SDDC or separate SDDCs can be used among tenants depending on available resources.
With vRealize Automation 8.2, provider administrators got the ability to share infrastructure by creating and assigning Virtual Private Zones (VPZ) to tenant organizations.
Think of VPZs as a kind of container of infrastructure capacity and services which can be defined and allocated to a Tenant. You can add unique or shared cloud accounts, with associated compute, flavors, images, storage, networking, and tags to each VPZ. Each component offers the same configuration options you would see for a standalone configuration.
vRealize Automation and VMware Cloud Foundation
With the pretty new multi-tenancy and VPZ capability a new consumption model on top of VCF can be built. You (provider) would map the Cloud Zones (compute resources on vSphere (or AWS for example)) to a VCF workload domain.
The provider sets these cloud zones up for their customers and provides dedicated or shared infrastructure backed by Cloud Foundation workload domains.
This combination would allow you to build an enterprise VPC construct (like AWS for example), a logically isolated section of your provider cloud.
SDDC Manager Integration and VMware Cloud Foundation (VCF) Cloud Account
Since the vRA 8.2 release customers are also able to configure a SDDC Manager integration and on-board workload domains as VMware Cloud Foundation cloud accounts into the VMware Cloud Assembly service.
VMware Cloud Director or vRealize Automation?
You wonder if vRealize Automation could replace existing vCD installations? Or if both cloud management platforms can do the same?
I can assure you, that you can provide a self-service provisioning experience with both solutions and that you can provide any technology or cloud service “as a service”. Both have in common to be backed by Cloud Foundation, have some form of integration (vRA) and can be built by a VMware Validated Design (VVD).
vCD is known to be a service provider solution, where vRA is more common in enterprise environments. VMware has VCPP partners, that use Cloud Director for their external customers and vRealize Automation for their internal IT and customers.
If you are looking for a “cloud broker” and Infrastructure as Code (IaC), because you also want to provision workloads on AWS, Azure or GCP as well, then vRealize Automation is the better solution since vCD doesn’t offer this deep integration and these deployment options yet.
Depending on your multi-tenant needs and if you for example only have chosen vCD in the past, because of the OrgVDC and resource pooling feature, vRealize Automation would be enough and could replace vCD in this case.
It is also very important to understand how your current customer onboarding process and operational model look like:
How do you want to create a new tenant?
How do you want to onboard/migrate existing customer workloads to your provider infrastructure?
Do you need versioning of deployments or templates?
Do customers require access to the virtual infrastructure (e.g. vCenter or OrgVDC) or do you just provide SaaS or PaaS?
Do customers need a VPN or hybrid cloud extension into your provider cloud?
How would you onboard non-vSphere customers (Hyper-V, KVM) to your vSphere-based cloud?
Does your customer rely on other clouds like AWS or Azure?
How do you do billing for your vSphere-based cloud or multi-cloud environment?
What is your Kubernetes/container strategy?
And 100 other things 😉
There are so many factors and criteria to talk about, which would influence such a decision. There is no right or wrong answer to the question, if it should be VMware Cloud Director or vRealize Automation. Use what makes sense.
I am a Senior Solution Architect at VMware and part of the solution engineering team in Switzerland. I engage with the VMware community and some of the largest customers in Switzerland. Opinions are my own.
VMware CMTY Podcast #546 – VMware Carbon Black Cloud Workload