Enterprise computing now occupies a new space called the “Hybrid Datacenter.” As organizations work to achieve the promises of cloud computing while retaining the benefits of on-premises deployments, they are increasingly operating in a new, highly heterogeneous environment that creates new challenges, including application service delivery. Application service delivery in the Hybrid Datacenter is one of the biggest challenges that IT Operations teams will need to meet for their organizations to be successful.
I like to start with a basic premise: IT organizations exist to deliver application services. Every business and organization is now highly dependent on applications. Whether it’s consumer applications for banking, scheduling, buying airline tickets, or coordinating youth sports team schedules, or business applications for HR, ordering, or production management; organizations of all sizes today rely on applications to function.
IT teams exist to make these applications work. Any change to the infrastructure used to deliver applications services – in this case the movement from on-premises to Cloud based computing – should be seen as an attempt to deliver application services with higher reliability, better performance, and lower cost.
The term “The Datacenter” is shorthand for both the physical building where the racks of servers live, and also for the overall enterprise computing environment. “The Cloud” is shorthand for public and private cloud environments. Now we’re seeing a new term that addresses the current state of enterprise computing – the “Hybrid Datacenter.”
What is the “Hybrid Datacenter?” (And weren’t we all supposed to have moved every computing resource to the public cloud by now?) Well, it turns out the IT market moves at its own pace. Cloud computing promises tremendous benefits – flexibility, scalability, and the ability to reduce capital and operational costs. These come with a perceived risk to the benefits of maintaining on-premises datacenter environment, which are seen as providing security and stability.
The physical datacenter hasn’t gone away. While many startups will tout being entirely cloud-driven, most enterprise-scale organizations still own most of their own servers. Even the most forward-looking organizations that are aggressively moving whatever they can to the cloud are retaining some in-house infrastructure.
The “Hybrid Datacenter” is the result of organizations’ efforts to combine the best of both of these options. When we work with enterprise customers, we typically see a combination of several of the following attributes:
Extensive virtualization within the datacenter
Use of multiple operating systems – typically Linux and Windows, but with Solaris and AIX thrown in for specialty operations
The beginnings of the use of containers – Docker as well as many other flavors
Multiple types of databases – typically SQL Server and Oracle, but some open source options as well
Third-party web services and SaaS offerings tied directly into the function of distributed applications
In terms of flexibility and deployment options, the choices have never been better. Developers have many options for testing and deploying new applications and services. But for the IT Operations teams who are responsible for ensuring application service delivery, things have become more complicated.
Most infrastructure management tools and systems do only that – manage infrastructure. Admin tools designed to report and manage systems based on availability metrics (i.e. CPU utilization, memory usage, etc.) are of limited use when trying to ensure the proper functioning of an application that spans tiers of web servers, application servers, middleware, and databases that reside in different locations.
To manage application service delivery in the Hybrid Datacenter, IT Operations teams need a comprehensive view of application performance. Specifically, they need:
Dynamic mapping and updating of each application’s infrastructure
Seamless visibility across all the components that make up each application – regardless of operating system, physical form, or location
Real-time response time measurements for both the application as a whole and for each application component
Real-time reporting to application stakeholders, including executives, operations teams, application owners, and development/QA teams
Organizations that implement solutions that deliver these capabilities will help IT Operations teams better solve the problem of managing applications in the Hybrid Datacenter.