FlexPod

Compute, network, and storage in an easy to configure, procure, and deploy model, you say? Today we are accustomed to the notion of grouping compute, network, and storage in a pod or stack. However, as we discussed in FlexPod Survival Guide: The Start of an IT Revolution, converged infrastructure wasn’t well defined when the FlexPod® solution came to market. Hyper converged wasn’t available either, and many organizations were just beginning to develop their cloud strategy. The emergence of converged infrastructure and FlexPod was influenced by multiple factors. Best practices, customer demand, partner readiness, and strategic alliances all had a hand in creating the market opportunity. Today, with many organizations adopting a hybrid cloud model, the engineering team views FlexPod as the foundation of the NetApp Data Fabric as we work on next-generation data center designs.

 

Since this is a bit of a FlexPod history blog, let’s share a few more stories. FlexPod is the only solution I have worked on in which the original name actually became the go-to-market name. So one might ask, “What is in a name?” As it turns out, quite a bit. Consider two very disparate companies, Pepsi and Google, originally referred to as “Brad’s Drink” and “BackRub.com”, respectively. In the early 1900’s, many considered sodas or colas to have medicinal qualities. Pepsi-Cola, derived from “dyspepsia” and the kola nuts used in the recipe, communicated that the drink was both a natural product and helped with indigestion. Google is a play on “googol” or 10 to the power of 100 because it is capable of searching and indexing an almost incomprehensible number of web pages. As mentioned in the previous blog post, when FlexPod was created, customers needed a standard infrastructure that could be tailored to their specific applications. Customers desired repeatability and prescriptive guidance while also requiring flexibility. Because we referred to our right-sized infrastructures as “pods,” the term FlexPod was used early on as it conveyed those attributes.

 

The team developed the core FlexPod principles by considering the foundational pillars of another joint reference architecture called Secure Multitenancy. The pillars were:

  • Service availability
  • Secure separation of workloads
  • Service assurance and the ability to deliver SLAs across the compute, network, and storage
  • Simplified end-to-end manageability

These pillars served as the team’s compass as creative ideas were developed and tackled. To ensure that no pillar was hindered by an architectural or product change, Cisco and NetApp developed standardized, scalable reference architectures for a variety of applications, workloads, and data center designs.

 

A common theme when FlexPod was first developed was a need for enterprises to deploy “a data center in a rack” housing various business-critical applications. This need led to the notion of a balanced reference architecture based on the number of virtual machines (VMs) and users of those VMs. It was an interesting concept, but over the years, the market has dictated that consumers need more flexibility for both on-premises and hybrid-cloud-based architectures.

 

Looking back on more than 150 validations the joint teams have developed, you will see a pattern of core infrastructure solutions and built-on offerings. We have validated a variety of hypervisors, application-focused solutions, and cloud-ready solutions. Each validation adopted emerging technologies and trends while reducing customer risk for new solutions. FlexPod, fully integrated with the NetApp Data Fabric, has been validated with all the existing hypervisors, the emerging container, and the DevOps space, and improves orchestration and lifecycle management.

 

One of the most recent additions to the converged systems portfolio is FlexPod SF. FlexPod SF takes advantage of all the capabilities of NetApp SolidFire® and the Element® OS for those customers looking to deploy a next-generation data center. The SolidFire Element OS offers unparalleled features for today’s next-generation data center and DevOps-focused scenarios.

 

FlexPodSFSome of the benefits of SolidFire include:

  • Advanced scale-out. SolidFire offers both linear and nonlinear scaling of resources. Scaling is nondisruptive and self-balanced.
  • Guaranteed performance. SolidFire provides fine-grain quality-of-service (QoS) settings that are without comparison. With SolidFire, administrators can allocate performance independent of capacity.
  • Global storage efficiencies. Deduplication is clusterwide with SolidFire nodes. Also, compression is performed inline.
  • Automated management. SolidFire offers a full-featured REST-based API. The storage is 100% programmable.
  • Data assurance. The SolidFire architecture affords RAID-less data protection. SolidFire clusters are self-healing and recover from disk or node failures in minutes.

Beyond a typical validation, FlexPod SF also introduces a new hardware platform in the form of the SF9608 storage node. The SF9608 node is an all-flash, scale-out, block-based storage platform built using Cisco UCS C220 M4 rack server technology and NetApp SolidFire Element OS software. Be on the lookout for future validations with this platform.

 

Over the past six years, customer designs have changed from architectures with only an on-premises focus to near-cloud, and even to solutions residing completely in the cloud. Cloud, like virtualization, is a disrupting technology that introduces different design considerations. The great thing about FlexPod is that Cisco and NetApp recognize that customers are on a journey and need to adopt supported technologies and features as their businesses allow. Put simply, no other converged infrastructure has the data management integrations and cloud vision that FlexPod does.

 

Never before has there been so much need for guidance in a hybrid cloud world through reference architectures. Exciting things are coming. FlexPod as the foundation, with Data Fabric as the vision, are driving principles for the team as we evolve and continue to solve customer challenges. It is exciting to think about what the next six years will bring.

 

For a deeper dive into recent history, check out the latest FlexPod NVAs and CVDs. Learn more about the NetApp Data Fabric, the performance, agility, and value of FlexPod SF, and the collaboration between Cisco, NetApp, and VMware on the original Secure Multitenancy. Finally, just for fun, visit the Pepsi and Google history pages to learn more about their origins.

Chris Reno

Chris Reno is a technical marketing engineer at NetApp in the Infrastructure and Cloud Engineering team. Chris is focused on architecting, validating, documenting, and evangelizing solutions based on NetApp products; specifically virtualization, cloud and converged infrastructure. Chris co-developed FlexPod and has held a variety of senior technical roles including sys admin, professional services engineer, and presales engineer. Most recently Chris was a presales engineer for a NetApp channel partner before being rejoining NetApp in 2017. In total, Chris has over 12 years experience with NetApp storage systems and has extensive experience in Q&A automation, virtualization, stateless computing, flash technologies, and hybrid cloud architectures.

Add comment