Interlocked

VMware has provided classic on-premises datacenter solutions with solid performance and support through ESXi, vSphere, and vCenter. With the rise of public clouds, to keep up with the times, VMware added products such as VMware NSX, an Open vSwitch extension, and the OpenStack compatibility layer.

About a decade ago, VMware also added a whole suite of tools under the vRealize label that enables the operation of hybrid workloads in public clouds and private virtualization environments. Since its debut, vRealize Suite [1] has matured, and version 8.1 was released in 2019. In this article, I discuss the components of vRealize and question their usefulness for you in everyday life.

VMware Before the Cloud

VMware has been making a name for itself in IT for decades as a provider of comprehensive virtualization solutions. Although VMware products often come with lavish price tags, the vendor delivers solid products for the cost. Anyone who has ever worked with ESXi and vSphere or vCenter knows that a well-designed interface provides the most frequently used functions in the context of classic virtualization, such as migrating a system from one host to another.

VMware is in complete command of these actions: live migration, high availability, and fully automated failover when a server goes down. None of these tasks present a major hurdle for VMware. Reliability is one of the reasons many administrators hardly bother looking at products by other vendors during their entire IT career. True to the saying, "Nobody ever got fired for buying VMware," you buy what you know and are happy that it works.

Then Came the Cloud

Anyone working in the IT environment today has experienced one of the biggest upheavals the industry has ever seen: the shift from small, local setups to large public clouds.

Admins and controllers have long since come to realize that operating their own IT infrastructure is expensive. It ties up personnel, by far the most expensive resource of all, but it also ties up money in the form of hardware and infrastructure costs for cooling, electricity (including emergency power generators), and other components.

In short, the idea of throwing away your own hardware and outsourcing the associated overhead to a large provider such as Amazon or Microsoft is meeting with enthusiasm in many companies. This business is also worthwhile for the providers. Because Amazon Web Services (AWS) and its partners have optimized all internal processes for the operation of such setups, they can offer this service particularly efficiently. Rumor has it that Amazon makes 97 cents in profit on every dollar in sales.

Customers are flocking to the providers in droves, despite the sometimes hefty prices, which shows how deep the sting of the costs of IT infrastructure operation is in the flesh of many companies.

Danger for VMware?

VMware is naturally not comfortable with this development, because it jeopardizes the company's traditional business model. If you no longer operate your own IT infrastructure, you no longer need on-site virtualization but rely instead on functions offered by cloud providers.

Many companies even leverage the fact that clouds do not offer virtualization with failover and all the other classic fuss as an opportunity to ditch old habits. When switching to the public cloud, the application per se is also put to the test. If benefits are to be gained from partial or complete porting of components to the cloud, you take this path rather than continuing to deal with operations yourself.

VMware therefore began some time ago to adopt a multi-pronged tactic in view of the threat of loss of importance. On the one hand, the company now has several technologies in its portfolio that extend typical virtualization with VMware to the cloud. Specifically, VMware NSX; an Open vSwitch extension that has matured into a comprehensive, highly complex solution for software-defined networking (SDN); and VMware's OpenStack compatibility layer, which allows vCenter environments to be controlled on the basis of OpenStack APIs.

VMware is clearly looking to capture revenue from those customers who don't see themselves as customers of AWS or Azure, but as their competitors – either private or public clouds. After all, there are more than enough end customers who are not allowed to store their data on AWS or Azure servers for legal reasons or because of their own compliance regulations. VMware is helping potential platform vendors who have these same customers with its own cloud portfolio.

The number of companies that see themselves as platform operators is likely to be small, however: The usual level of turnover cannot be maintained with these companies. VMware is therefore also approaching customers directly who want to use services such as AWS or Azure to some extent but still want to operate their own infrastructure in parallel (e.g., for critical data).

The VMware solution to this problem is vRealize, which enables the operation of private, hybrid, and multicloud environments. Of course, the VMware infrastructure is at the heart of this; the solution is primarily aimed at customers who already run VMware in their datacenters. The solution also results in a migration path: If a company wants to reduce its own infrastructure, VMware vRealize will be the tool to help it do so.

Four-Component Adhesive

In the hybrid cloud context of VMware vRealize, the company promises nothing less than a panacea. According to the manufacturer, you do not have to take care of anything in a vRealize environment. VMware promises that the deployment of any workload across virtually any platform is just as easy as the use of certain cloud APIs, such as AWS, Azure, or OpenStack. Those who already have VMware products can just as easily deploy and link them directly to vRealize.

However, vRealize doesn't just take care of the deployment, it also takes care of monitoring, alerting, and trending on the fly, automatically scaling workloads across the board. If your environment runs out of steam, vRealize adds resources to suit your requirements. To keep you informed about what the environment is doing, vRealize also includes a solution for centralized logging.

According to the manufacturer, vRealize meets compliance requirements by connecting to and supporting existing processes such as ITIL, a set of best practices for IT service management. If the vRealize scope is not enough, the extension APIs are a great choice. Not only are they openly designed, but they also let you retrofit arbitrary functions – if you are willing to put in the time or money.

Reading VMware's marketing brochure about vRealize, you quickly gain the impression that you don't have to do anything except roll out a setup, which is of course not true. vRealize does not relieve you of all the work, but it does shift some of the load off your shoulders in various areas.

Automating Features

The key feature in the tool suite is vRealize Automation (Figure 1), which is basically a large orchestration framework that calls functions in different cloud environments under a uniform interface (and, if so desired, behind uniform APIs) (Figure 2). What sounds abstract in theory, is quickly intelligible in practice.

Any administrator who has ever dealt with different cloud environments knows that the same features are available in one form or another from virtually every vendor. Creating a virtual machine (VM) for a specific purpose bears witness to this. To begin, you usually create a virtual network, which is then connected to a physical, external network with a virtual router. Then, you create a storage volume for the VM, usually based on an existing operating system image. Finally, you define the SSH key that needs to be active on the VM and start the VM itself. A short time later, it is ready for use.

The problem is that with hybrid workloads spanning multiple public clouds or a private and public cloud, you have to do this work multiple times. Moreover, you have to take care of some integration work to keep the setups compatible with each other. For example, if you run OpenStack in your datacenter and want to run a shared setup between OpenStack and AWS, you need several templates in the different languages of the orchestrators. This process is tedious and challenging during operation.

Automation Takes the Helm

In such a scenario, vRealize Automation comes in to play. The product is designed to communicate with the APIs of various public clouds, as well as those of private virtualization setups. Logically, vRealize Automation can also dock onto existing vCenter setups. However, OpenStack is also on the list of supported environments in private datacenters.

Instead of building separate templates for each target environment, you use vRealize to build blueprints and service descriptions of standardized services. Automation then ensures in the background that these are rolled out on the desired target system when called by the user. vRealize Automation is primarily an abstraction layer for other cloud services to create a uniform interface in either graphical or API form.

VMware's understanding of its core target audience is quickly reflected in the application: Preferably, users should use a self-service portal where they can quickly click together the desired workload. Additionally, vRealize Automation integrates comprehensive governance features. The admin first stores the credentials for the target clouds, but the product also comes with its own rights management, which lets you fine-tune who can do what down to the level of individual users. The same is true for quotas, which you also manage centrally in vRealize Automation.

As usual, VMware does not compromise on the supported target setup: The manufacturer supports generic clouds such as AWS and Azure, as well as its own products, such as the SDN software NSX.

Deployments and Operation

Modern cloud-ready applications and, especially, the container-based apps of today have different operational requirements than conventional applications from the distant past. Traditional monitoring, for example, must be much more flexible in clouds than in legacy environments, because the number of instances to be monitored can change abruptly and dynamically.

For reasons of scaling alone, workloads in clouds live because they specifically book and use only those services they actually need at time t. An online store for fishing equipment, for example, will need a lot of resources during the fishing season in summer but will have less load in winter. Modern web applications are therefore built in such a way that they can start any number of new instances of themselves on demand to distribute workloads across more shoulders.

The administrator rightly expects that a hybrid workload management solution will take these factors into account and address them accordingly. vRealize does not let you down at this point: vRealize Operations and Log Insight make sure you have a clear view.

Scale, Monitor, Reconfigure

The Operations component takes care of all tasks that arise in everyday life after the initial deployment of a setup. If it detects that the environment is running too slowly by referring to, for example, preset performance values, it automatically creates virtual instances. This works for conventional VMs as well as for containers, which the entire vRealize suite can handle with ease. vRealize Operations also takes care of capacity management. If RAM and CPU usage is too high for a given instance, it automatically restarts the instance.

At the same time, Operations dynamically adapts the configuration of workloads in cloud environments. After a scaling process, if it is necessary to adjust individual values in the configuration files, you can so specify under the Blueprints tab (Figure 3) for the respective environment. vRealize Operations observes such instructions in its templates and makes the appropriate changes, if necessary.

The monitoring aspect is not neglected either: vRealize Operations provides a cornucopia of functions to monitor target instances thoroughly in deployments, ranging from simple tests, such as whether an instance is responding to ICMP requests (ping), to complex checks that you can integrate into a vRealize blueprint.

Special Topic Logfiles

VMware even dedicates a separate component in vRealize to logfiles, which is more complex in distributed environments than in conventional environments because you have to deal with more components. Not all environments store their log messages in files on disk. For example, if the Podman container solution is used, it keeps a separate logfile for each container with the output of the standard output channel stdout and the standard error channel stderr. If a setup is distributed over several cloud instances, the complexity increases considerably. In the worst-case scenario, you then use different cloud environments with different techniques to find a problem.

Here, vRealize enters the scene with its Log Insights component, which aggregates the log messages and files of a setup if you have previously specified them in the vRealize Automate blueprint. The structure of the deployment is then no longer important – vRealize instead collects all relevant logfiles in a central location and allows quick access (Figure 4).

In summary, vRealize Operations and vRealize Log Insight are the extended arms of vRealize Automation. The two components monitor what the virtual setups were assigned at the outset and ensure that all the rules you set are also fulfilled.

The issue of compliance also comes in. Ideally, you will want to take this into account when configuring a blueprint for an application in vRealize Automation. In this way, you can ensure that the mandatory relevant rules are applied according to the company's compliance policy. Manual changes to setups can, at the very least, result in virtual setups no longer meeting compliance requirements. vRealize Operations puts a stop to this by constantly comparing the status of virtual setups with the target and by revising manual changes where necessary.

vRealize Business

Running setups and keeping them in line with compliance requirements is good, but in most cases, companies also want a certain form of integration into existing company processes.

On-site admins might be tempted to roll out a virtual test instance on AWS that occasionally is forgotten and not deleted afterward. A few cents a minute doesn't sound like a lot of money, but if a virtual instance remains unused for a whole month, it can cost into the thousands of dollars. An admin that uses vRealize will be able to see the costs of the different clouds when starting workloads, and an automatism will switch off instances that run too long according to defined rules.

Such features are implemented by vRealize Business for Cloud. Its most important tool is a central dashboard that allows you to keep track of costs already incurred and those still to be incurred. If desired, categorization can take place here as well. Different users and the costs they generate can be grouped, so you can quickly identify the projects that generate particularly high costs.

The stated goal of vRealize is to make costs across different types of infrastructure transparent and comparable for controllers. When choosing a target workload platform, you not only have the opportunity to make a decision in the best (i.e., most cost-effective) interest of the company, you also get a concrete feeling for which infrastructure causes what kind of costs.

The prerequisite, however, is that someone enters the costs incurred for parameters such as vCPUs and vRAM in the tool for their own use case. Depending on the contract with the cloud service providers, these costs may well differ from case to case (Figure 5).

vRealize Business also has practical added value for administrators of private clouds. If you grant admin rights to the user who connects to a private cloud with vRealize Business, the tool reads the current usage statistics from there.

vRealize then implements a trending function. If the hardware is no longer sufficient for the expected workload in time period t, the software sounds an alert and prompts you to add resources. This function is advantageous in private clouds because, procuring additional hardware requires some lead time. Thus, vRealize helps with planning.

Manager, Manage Thyself

Regardless of all external factors, vRealize includes a fifth component that functions as a kind of metacomponent. The individual vRealize components somehow have to find their way into a customer's infrastructure. The vRealize Lifecycle Manager takes care of exactly that by rolling out the aforementioned components and integrating them with core VMware tools, such as VMware Marketplace. From there, prebuilt application blueprints can be integrated directly into vRealize tools, such as vRealize Automator.

Once again, VMware proves that it is capable of more than just providing reliable life-cycle management for its own products: VMware also regularly succeeds in building functional ecosystems around these tools, leaving you largely happy.

Beware of Lock-In

Once you've aligned your internal workflows with the VMware toolchain, you can't get away from it – at least not without significant investment. Migration of existing vRealize setups to other tools is virtually impossible. If the close relationship with VMware is okay with you, it's not a problem; many VMware customers know the principle anyway. Nevertheless, you should consider this effect in advance.

What Does It Cost?

From my point of view, the topic of license costs at VMware is regularly less appealing than the technology. The vendor has a very strongly fragmented price strategy, with Standard, Advanced, and Enterprise editions of the products.

VMware itself only gives some nebulous hints on its product homepage with regard to the feature set on offer. For example, the Enterprise Edition offers all features that are basically available in vRealize, whereas the Standard Edition only offers basic functionality. However, admins are unlikely to be able to make complete sense of the categories. What's more, a direct feature comparison of each edition, allegedly available on the VMware website, was a link to a black hole at press time.

Ultimately, as is common with VMware, only a chat with a VMware distributor will lead to an understanding of prices. The distributors can deliver a concrete offer with actual figures, which you traditionally cannot get without contacting their sales force. However, VMware explicitly states that vRealize can be used as usual with a multicloud license from the company.

Conclusions

VMware vRealize is a typical VMware solution: It is easy to install and run, and its rich feature set delivers a happily-ever-after solution. If you already have VMware in-house and can get it cheaply, consider VMware vRealize as a solution for running hybrid setups. The feature-rich capabilities do what the vendor promises and feel as you would expect with VMware.

However, look before you leap: Once you have rolled out vRealize setups, switching to a different solution is painful. The lock-in effect hits home hard and makes it difficult to escape later.