Democratizing Cloud Usage with Digital Intelligence for Multicloud

blog image
Target Audience: Users of NetApp Kubernetes Services
 
Desired Action After Reading Post: Sign up for free trials.
 
Topic – How NetApp Kubernetes Services users benefit from using Fderator.ai from ProphetStor
 
Keywords: NetApp NKS, Kubernetes Management, MultiCloud, Day 1 Operation, Day 2 Proactive Management helps to boost performance and reduce cost in moving to Public CloudsOperation
 
Outline
Proactive Management helps to boost performance and reduce cost in moving to Public Clouds
 
Microservices, Kubernetes, and MultiClouds are winning.  They also create issues that need to be addressed for universal acceptance.
Cloud helps free the developers and data scientists from operating in the legacy IT infrastructure so they can focus on developing quickly, deploying easily, and adding values to differentiate the company from others.  Life was easier in the days when there is only one cloud service provider.  Nowadays, the developers need to release on other public clouds, dealing with the difference among clouds, and it becomes a big hassle.  Microservices are wining, and Kubernetes is wining.  However, as putting together release strategies, building microservices, and support Kubernetes services on MultiClouds becomes a norm, the developers are joggling with the deployment, and they need a service, rather than expensive software, to simplify the task.
Anthony Lye of NetApp once said, “We Loved Stackpoint.io So Much, We Had to Buy the Company!”  StackPointCloud takes the complexity out of the deployment of Kubernetes to MultiClouds, including the upcoming NetApp HCI private cloud offering.  We consider this a great solution to the Day 1 operation of moving to the Cloud.
StackPointCloud is now NetApp Kubernetes Services (NKS).  It helps the developer to do more development than deployment.  Besides, NKS later becomes a platform that helps developers agnostically manage application life cycles on different clouds, Helm charts, and K8s services (AKS, GKS, EKS).  With the provision of the certified Cloud Native Solutions, the user can now extend the functions of the cluster in a way similar to you adding Apps to your iPhones to customize and enhance the user experiences.
This platform can utilize Machine Learning technology to the Operations of the Kubernetes and containerized applications in the multicloud.  Although we are enjoying the benefits of microservice-based applications when comparing to the original, monolithic applications, the complexity of managing the communications among the pods in the clusters deployed is becoming prohibitively high.  Providing the specifications for the scalability of the pods and clusters by human input offsets the benefits of the autoscaling. Worst yet, the conventional wisdom inherited from the reactive management, taking action only to handle unplanned outage or unexpected performance variations, increases cost and impede the performance.
Together with NKS, Federator.ai Boosts Performance, Reduces Cost of Operation, and Enables Planning
Federator.ai for NKS is a Trusted Chart cloud-native application designed to change the status quo.  It collects the operation data from multi-layers, including the application layer, Kubernetes Layer, and Infrastructure Layer, to name a few, so that correlation among layers of operations can be identified.  Then the Data Process Engine turns the collected data into insights, creates foresight predictions for operations, including workload predictions for containers and namespaces.  With the prediction, we now have a way to create recommendations for the spec for auto-scaling.  The native Kubernetes autoscaling handles the scale-up and scale-down of the size of the pods or number of replicas of the pods only when the CURRENT condition meets preset criteria.  It is REACTIVE and incurs latency to act.On the other hand, if we know upfront what will be needed, the recommended resource adjustments can be performed “Just-in-Time-Fitted” in a PROACTIVE fashion, and we can expect to see the performance improved as the latency is reduced.  In many cases, we have seen latency reduced anywhere between 45% to 75% for applications such as NGINX and Kafka!  By providing the recommendation, customers now have the clarity about what Request/Limit they have to set for pods/namespaces.  Unwanted waste on overprovisioning no more.  This will equally please the developer and the executives paying the cloud service bills!
Federator.ai also provides the transparency of the costs running a workload on different public cloud service providers.  Not only it provides the cost for the customer-specified cloud service providers, but it also lists the top 3 choices, providing opportunity for economic arbitrage for running a workload.  With “cost of operation” becoming the top consideration for moving to the cloud, the Federator.ai provides the means to democratize cloud services.  Together with NKS, we are providing the users with the freedom to run the workload on any cloud, with the desired cost and SLA.

Day 2 Operation is Essential for the Ultimate Success of Moving to the MultiCloud

Federator.ai brings customer bonding to NKS as a platform.  Together with NKS, we are helping the customers to handle Day 2 operations that include
  • Monitoring (for performance and issues)
  • Resource Orchestration (where to put the pods/clusters on nodes, considering their communication, workloads, and duration of services)
  • Automation with Optimization (continuous engagement with the deployment and autoscaling recommendation) — Point Decision Making to Foresight Optimization
  • Performance Enhancement without code change
  • Just in Time Fitted based on ML-Digital Intelligence
  • Economic Arbitrage (Demand Side Platform)
  • Issue Isolation and Resolution (MultiLayer correlation for MultiCloud)
  • Fast Fixes Validation
Currently, Federator.ai is available as a free trial and can be installed with one-click through the Trusted Chart from the web site of NKS.  The user can access the recommendation through REST APIs, grpc calls, CRDs, or the APIs for the InfluxDB that is the repository for the recommendation for scaling. We believe the NKS platform and Federator.ai for NKS together will much reduce the hurdles for people to move to MultiCloud in the era of Cloud 2.0

Please select the software you would like a demo of:

Federator.ai GPU Booster ®

Maximizing GPU utilization for AI workloads and doubling your server’s training capacity

Federator.ai ®

Simplifying complexity and continuously optimizing cloud costs and performance