The Covid-19 pandemic has highlighted just how critical it is for businesses to be able to manage infrastructure, tools, and processes remotely and how important it is for infrastructure to be virtualized and digitized. Unfortunately, in many cases, IT management was not prepared to support mass remote access to the technologies and tools that workers needed to perform their jobs. Many enterprises lacked the agility needed to react quickly to the rapid increase in remote connections due to poor capacity management and insufficient documentation. As a result, there was an enormous amount of stress put on networks that did not have the bandwidth to handle the changing demands.
To gain a better understanding of how infrastructure can support changing requirements, such as widespread remote work, let’s take a closer look at how the management of data centers, cable networks, IT assets, and edge sites is evolving. Traditional data center infrastructure management (DCIM), which ensures that energy, equipment, and floor space are optimized, cannot manage today’s modern data centers. They are just too complex and require new methods.
Modern data centers require a different approach
Today’s organizations have a hybrid digital infrastructure consisting of both physical and virtual assets and functions, with physical and logical resources, connections, and dependencies. Part of the infrastructure is on premise, and part is in the cloud. Though some assets and resources reside within the data center, more and more are extended to telco and IT.
Conversely, many resources not typically part of the data center have migrated there. Telco resources are a good example of this. As they become virtualized and no longer part of the physical communications network, they by default are moved into the data center. Either way you look at it, managing today’s hybrid digital infrastructure requires hybrid digital infrastructure management (HDIM).
HDIM extends the functionality of DCIM. DCIM is focused on space, power, and cooling. HDIM includes those as well as everything digital and the network. Businesses need solutions that encompass all to manage their hybrid digital infrastructure most effectively.
Unified resource management is key
Given today’s geographically distributed IT resources, IT departments need to evaluate, manage, and optimize the entire infrastructure inclusive of the central data center, individual edge locations, and their connection to the main data center. Diverse resources across the building infrastructure (power, cooling, floor space), the IT infrastructure (networks, servers, storage), connectivity (physical cabling infrastructure and logical circuits/bandwidth), and services (software, applications) need to me managed holistically.
Insight into all physical and virtual assets and their dependencies, manufacturer-independent and neutral, can only be achieved through unified resource management with a uniform data model. Hybrid digital infrastructure management starts with an integrated inventory of all physical and virtual assets, and logical connections and dependencies, contained within a single repository that is accessible to all users throughout the organization. This data is the foundation for managing the hybrid infrastructure core.
Some solutions also provide 3D representations and simulations to help IT managers visualize the information stored within the central database and simulate changes. A dynamically updating database can ensure data consistency and accuracy, which is critically important for planning, operation, and fulfilment teams that rely on that information to make business decisions. This approach ensures the greatest level of data accuracy for the upcoming changes.
Best practices for choosing a HDIM tool
The following domains clearly illustrate the differences between DCIM and HDIM tools, yet are often overlooked when selecting data center management software:
1) Cable management evolving into connectivity management
Traditionally, patch management, outside plant management, and end-to-end connectivity are considered three separate domains and are managed by separate teams using separate tools. Today, due to digitalization and virtualization, the same resources need to be leveraged across all three domains. Assets and services in modern complex networks must be managed within a unified management tool regardless of where they live within the organization for complete transparency throughout the infrastructure.
Conventional DCIM solutions often do not include cable, patching, or network management functions at all. Instead, they focus predominantly on space, power, and cooling, therefore separating the network from all other data center domains. This segregation has shown to be inefficient and problematic for both planning and operations.
Network connections are an integral part of all move, add, and change processes and are also needed in all maintenance activities. IT without network is virtually useless at any power redundancy level so “removing” this aspect from the main software tool for data center management is creating a dangerous blind spot. All network related resources including panels, patch cords, backbone cabling, trays, etc. need to be managed with the same quality as the power chain or active components such as servers and switches.
2) Digital twin
Digital transformation within data centers and colocation sites is changing IT layouts and increasing their complexity. To stay on top of this, operators must create a digital twin, or a virtual replica, of the infrastructure. The digital twin can support all potential and actual physical assets, networks, and processes to provide clearer insights into all data center locations. The pairing of the virtual and physical worlds within a live model allows operators to analyze data and monitor systems to optimize space management and mitigate problems before they even occur. A digital twin will also enable operators to prevent downtime, develop new business opportunities, and accurately plan for the future by using simulations.
For full transparency into all of the components across the organization, a digital twin needs to model on-premise equipment as well as all resources in collocation and the cloud, core and edge sites, and all IT, network, and facility rooms. All elements matter and their documentation need to be harmonized across types and locations.
3) Single pane of glass for capacity management
Hybrid is not only bringing new complexities to operations and planning, but also to capacity management and forecasting. Bridging the gap between the various domains that operate the elements of your hybrid infrastructure is crucial to obtain valid data for these processes. As not all systems and subsystems from facility, IT, and networks are managed within one umbrella system, you still need to assure that a system of record is capable of providing consolidated and normalized data from all domains to properly manage your capacities across all sites and technologies.
Using a common set of validated data for all your operations and planning processes will improve efficiency and quality as all teams can leverage the same reliable information and constantly update and validate existing documentation automatically in the wake of their activities. This greatly reduces efforts to manually update documentation and redo planning as it is no longer based on outdated information.
If your organization has already implemented the concept of the digital twin, you can use that system as the single source of truth for your analytics processes to monitor and forecast capacity usage and demand.
4) Service-driven data center operations
An important aspect to consider for contemporary data center operations is the move to service-centric design thinking. Rather than simply focusing on individual, “unrelated” assets and technical components with physical redundancy in a conventional bottom-up design, data center operators must find ways to differentiate themselves in the marketplace through additional value-added services such as comprehensive SLA reporting, integrated current/power measurement, pay-per-use billing, and an extended connectivity portfolio within one facility or across all locations. This concept is a lot more suitable to tackle the challenges of hybrid environments.
In order to acquire the necessary data center resources and provide the services customers need, operators need to be able to plan their IT, network, and physical infrastructure in detail and maintain full transparency across all data center assets. To achieve maximum efficiency when delivering services, data center management procedures and tools should provide asset configuration and change management, documentation of asset associations and maintenance history, connectivity analysis and port usage, and the monitoring of power consumption on devices.