Seamless Growth in the Scale-Out Era
Change is inevitable and it may be a fortunate consequence of growth.
In an era where agility is needed and delivering new and better services faster may be the difference between outpacing our competitors or not, IT is many times perceived as one of the major roadblocks.
For example, think about a new Internet Connection Service needed at your new home. Would you choose a vendor that told you it takes them 6 weeks for the technician to arrive and activate the service or would you elect one that would do it within a few hours just by activating it remotely?
I bet the vast majority of us would choose the latter, while the former would need to eventually catch up to deliver such services to be at least considered by most us.
Although at the application development level we are moving from Waterfall to Agile methodologies to deliver new services, the infrastructure (whether it may be in the cloud or on-premise) is still playing a major role and it can potentially become a bottleneck for any transformational project.
Having per-cloud, per-location or per-technology siloed networks, server-farms and datastores is no longer an option and we demand more intelligence from our multicloud data center infrastructure to keep up with the ever-needed change to drive growth in any business.
If companies expand or go through mergers & acquisitions, they have to undergo a set of complex tasks and expensive carrier services to, for example, provide either LAN-to-LAN connectivity, buy on-prem to cloud services like AWS Direct Connect or CSR-1000v or keep the remote environment connection separated from the current networking addressing & management. In any case, local IT personnel at the remote location or for a specific cloud is often needed.
Back in 2014, the monthly cost of a 100Mbps EoMPLS pseudo-wire connection in Miami or Buenos Aires was $5321 and it has been declining at 30-40% yearly according to a study performed by TeleGeography, however although prices decrease, an additional budget is still considered for DCI and LAN to LAN services while the same operational complexities still happen.
The same occurs with Compute and Storage, where multiple servers may be deployed locally on every remote location to provide services like Active Directory or File Shares that people may want to access faster but that usually implies additional IT staffing as well as diverged management.
With solutions like ACI for Data Center Networking we can add today a new branch office with a fully-automated VXLAN deployment in only 2 minutes as shown in the video below. This is a major game-changer since this means $0 for DCI or LAN-to-LAN links plus no local IT staffing since everything is managed centrally. Moving forward with ACI Anywhere, we could also be able to even extend a single point of management and VXLAN connectivity to IBM Cloud or AWS with technologies like ACI vPod and ACI Multicloud.
The same applies for Compute and Hyperconverged Storage with UCS and HyperFlex, where we can manage and provision everything from Intersight and add remote-locations using HyperFlex Edge in just minutes as shown in the video.
Once those resources are added, there may be some configurations that may conflict with the main site ones. Sometimes we leave behind erroneous or unfinished configurations that may cause network or IT outages.
A study made by the Ponemon Institute in 2016 reported that 22% of all network outages were caused by human errors. What if we could minimize the risk associated to erroneous configuration or even know the impact of change in advance?
Today we can. With tools like Network Assurance Engine, risk may be reduced dramatically by providing 5000+ built-in failure scenarios in a platform that proactively analyzes all policies and configurations in your network.
With technologies like these I think we could make the popular phrase “Growing pains” have some exceptions!