your123 - stock.adobe.com
Automation, cloud lead as influential data center trends
Recent data center trends, such as AIOps and automation, have been geared toward simplifying network management. Here are the trends and challenges teams faced over the past year.
Since the beginning of 2020, IT departments have focused on migrating apps and data into the cloud, as well as bolstering remote access technologies to account for the massive uptick in remote workforces. One area that may have been neglected during this time is the private data center.
Now that the pandemic dust has largely settled, IT leaders are looking at their data centers in terms of modernization and integration in hybrid and multi-cloud infrastructures. Teams that let their data center ambitions lapse for a period of time might find it useful to read up on enterprise data center trends and challenges.
Data center trends in 2021
Recent data center trends tend to gravitate around the simplification of networks and reduction of manual tasks performed by IT staff. A few ways teams work to simplify network management include the following:
- simplifying cloud connectivity;
- automating network performance; and
- enabling AI for IT operations (AIOps).
Simplifying cloud connectivity
Distributed applications and workloads that extend from private data centers to clouds often require network architects to rethink connectivity to provide performance and security. This may mean a shift toward WAN or AWS Direct Connect links with dynamic routing in between. It can also mean shifting certain time-sensitive services from the cloud to metro edge computing counterparts.
Either way, the focus on networking in 2021 largely revolved around how to gain performance in data that flows to and from the data center, as opposed to within the data center.
Automating network performance
data center automation can mean many things, but from a data center trends perspective, network flow performance optimization is what's hot right now. This optimization is when AI built within the network's architecture combines with real-time streaming network telemetry analysis. The combination can intelligently determine the optimal path for a particular data flow at any given moment.
Another use of AI in the data center is streamlining manual processes when monitoring and troubleshooting data center networks. AIOps uses a host of network monitoring methods combined with advanced AI.
AIOps first creates a baseline of the data center for normal activity. It then provides intelligent insights when traffic flows in the data center veer off from that norm. These insights often include root cause analysis and steps to remediate network performance or data security-related incidents.
Data center challenges
Data center challenges, both new and old, still exist when enterprises attempt to upgrade private data centers with the latest technology components and architectures. Some data center challenges that network teams might run up against are the following:
- juggling legacy and modern network architectures;
- optimizing access and performance for remote workforces; and
- dealing with a shrinking pool of data center management.
Juggling legacy and modern network architectures
It's impossible to upgrade every network component in a data center. As a result, network teams sometimes have to grapple with managing modern leaf-spine architectures alongside legacy network components managed on a hop-by-hop basis. These types of challenges can prove difficult as shifts in software, server and service architectures continue.
Because of this challenge, data center architects and administrators must possess a unique mix of old and new data center skills as they eventually migrate data centers to centrally managed and software-defined networks.
Optimizing access and performance for remote workforces
Because workforces are so distributed these days, it's sometimes best for teams to redesign a data center to better optimize access from outside the corporate network. Many remote users still use VPNs to connect to a corporate LAN edge, only to traverse multiple hops before reaching the on-premises data center located in a different physical location.
Using modern secure remote access technologies can eliminate network hairpinning, which is the process of rerouting a packet back to its origin point before sending it to its destination. Removing network hairpinning can shift the access edge to a more optimal location. It can also enable the use of work-from-home network hardware or software agents to manage flow optimization between the client and server.
Dealing with a shrinking pool of data center management
Due to the explosion of public cloud computing, fewer IT professionals wish to build, manage and maintain a data center from bare-metal components on up.
Instead, these IT pros have largely shifted their skills focus toward the deployment and management of cloud computing instances. Because of this shift, it's becoming increasingly difficult to find talent who are willing to work in physical data centers, as opposed to virtual ones that reside in remote cloud networks.
What does all this mean?
All the multi-cloud connectivity, automated processes and AI will lead us down the path to data centers that require far fewer on-site IT operations staff to manage. Remote monitoring and management will replace on-site staff, and AI will eliminate many manual processes. What's left will be a data center infrastructure that is highly flexible, low maintenance and scalable.
Experts predict hot trends in cloud architecture, infrastructure
8 ways to make network modernization work