Contributed by Vaibhav Dongre, Business Head, EMEA & APAC, Robin.io.
I have been involved with this big idea of The Future Telco for the last several years and since the early days it was clear that networks of the future will have to behave like the human neural system – predicting, anticipating, continuously learning, adapting and autonomously self-correcting.
Now, as we get closer to the EDGE, there is increasing evidence of data and network intensive applications, leveraging advancements in Artificial Intelligence (AI) and Machine Learning (ML) technologies, demonstrating the exact same behavior – dynamically responding to the demands of millions of connected devices which are in-turn aiming to manage critical functions in day-to-day life pushing automation to scale. This behavior is similar to the autoimmune functions in the human neural system. Our Future Telco networks need an application lifecycle management system which is always ON, manages hundreds and thousands of applications in a fully automated, secure and scalable manner without any human intervention – a neural network, just like our human neural system.
Drawing on the parallels in the telecom world, open-source container-orchestration systems such as Kubernetes are gaining popularity for deployment and automation of Application Lifecycle Management (LCM). While Kubernetes is great for deployment, scaling and management of stateless applications, for complex data intensive and stateful applications such as Open RAN deployments, orchestration of CU & DU et al, Kubernetes has some inherent challenges such as,
- Always On – How can data be always available even if Pods, Disks and Servers fail?
- Simple – How DevOps teams can manage Stateful Apps without becoming storage experts?
- High Performance – Predictable app performance is critical in a shared cluster
- Business Continuity – Snapshots, Clones, Backup & DR are required for business continuity
- Observability – Observability is critical for maintainability and operations
For several promising 5G EDGE use cases such as remote operations, autonomous and hyper-connected devices, machine-to-machine communications for IoT and smart city use cases, cloud-native approach to LCM automation and infrastructure orchestration does seem to hold promise if these challenges are addressed.
INNOVATION AND ADVANCEMENTS TO CLOUD NATIVE (read KUBERNETES)
Kubernetes is rapidly becoming the de facto standard for deployment and LCM of cloud-native applications and thanks to the award-winning innovations in the space of persistent data management, the challenges associated with data management of stateful applications on Kubernetes have been addressed at the architectural levels. With the world’s first end-to-end cloud-native deployment of 5G network in the far east, these patented innovations are paving way for hyper convergence and scalable automation. These capabilities have now been proven in deployment and are ready to become the underlying fabric (cloud-native platform) for the Future Telco networks – dependable, secure, autonomous, self-healing & auto-scaling, responsive & predictable, delivering high performance for data hungry applications and enabling real-time decision making to unlock the full potential of ORAN and EDGE deployments.
SIMPLIFY & FUTURE PROOF CLOUD NATIVE FOR THE DEMANDING EDGE
While the EDGE promises to unlock a multitude of opportunities worth billions of dollars for operators and private enterprises, it is a high constraint and demanding environment. Imagine a few nodes at the edge location trying to feed into a bunch of data and network-intensive applications all contesting for compute and storage resources simultaneously while the applications themselves have to guarantee the highest performance and real time decision making to deliver the promises of low latency, high throughput data and network intensive 5G applications.
The underlying cloud-native fabric (platform) which is most ideally suited for the edge environment will be the one that offers carrier-grade networking, advanced compute and storage capabilities (hyperconverged platform), fully automates day 0/day 1 activities such as auto healing, auto scaling, snapshots, backups et al (Zero Touch Automation ZTP) and is a lightweight operator (application) built for managing very large scale operations right from bare-metal provisioning to LCM, and orchestrates workflows at scale across multi-cloud, multi-cluster environments and offers a single pane of glass management without leaving a significant footprint.
THE FUTURE TELCO APPROACH AT THE EDGE
While the EDGE promises to unlock large-scale potential opportunities across segments such as Enterprise 5G, Media & Entertainment, Industry 4.0, IoT, and Smart city, there has been little progress in the past due to various challenges. For example, the solutions for various industry verticals need the specific domain expertise to deliver an efficient end-to-end solution and most of these use cases need an ecosystem play beyond just connectivity and managed services.
The Future Telco (read progressive regional operator) has lately been engaging in bringing together various best of breed ecosystem partners such as application providers, infrastructure providers and cloud providers to bring to life the end-to-end use cases in the identified segments of their enterprise customer footprint. Various consortiums, test beds and production ready labs are on the prowl across regions that are actively trialing such solutions for monetisation of 5G network infrastructure. There is also an increasing interest in taking the success stories to other operator peers who then do not need to reinvent the wheel. Alongside the disaggregated technology stack, operators are now cherry picking the best players in each of the solution component categories to stich these solutions together.
ROBIN FOR THE EDGE
With over 72 patents in the cloud-native space and 3 award winning products Robin.io has brought several key innovations to Kubernetes to make it ideally suited for EDGE deployments. Robin’s industry leading Cloud Native Storage (CNS) is ranked as the most performant software defined storage and is purpose built for Kubernetes. Robin CNS Provides full automation for day 0/day1 operations and multi-cloud mobility.
Robin’s Cloud Native Platform (CNP) supports both VNFs and Containers within the same platform and brings the much desired hyperconverged capabilities with carrier grade networking and enterprise grade storage stack with advanced data management capabilities, LCM with single click deployment of complex application pipelines and with provision for CI/CD integration.
Robin has also built a highly scalable orchestration product called Multi-Datacentre Automation Platform (MDCAP) which orchestrates a variety of workloads including appliances such as baremetal, servers, switches, routers and up to a million elements.
Robin platform and orchestration solution has been one of the first commercial end-to-end cloud-native deployment of 4G/LTE and 5G workloads in production and at scale for a greenfield operator in the far east. Robin is presently engaged with several industry leading operators across regions for solutions related to deployment of ORAN and MEC use cases to accelerate the cloud native transformation, drive automation and simplify the lifecycle operations.