Hyperconverged Infrastructure the Evolution of Convergence Continues
• Defining Hyperconvergence
• Examples of Convergence
• Business Drivers for Adoption
• Competitors in the HCI Arena
• Impactful Future Technology
• Shifting Business Foresight
Forrester defineshyperconvergence:An approach to technology infrastructure that packages server, storage, and network functions into a modular unit and adds a software layer to discover, pool, and reconfigure assets across multiple units quickly and easily without the need for deep technology skills..
Hyperconverged Infrastructure is Characterized By:
• Commodity x86 hardware components that combine hypervisor, compute, storage, and storage switching with other IT services, such as data efficiency and data protection, in the stack, effectively eliminating the need for discrete IT components
• A single “building block” appliance that, when combined with additional building blocks, provides a single, scalable resource pool; and seamlessly scales in capacity and performance
• A high degree of automation
• The ability to manage aggregated resources as efficiently as possible within and across data centers as a single federated system and through a common toolset
The Seven Key Concepts Behind the hyperconverged Infrastructure
Elasticity: Hyperconvergence makes it easy to scale out/ in resources as required by business demands.
VM-‐centricity: A focus on the virtual machine (VM) or workload as the cornerstone of enterprise IT, with all supporting constructs revolving around individual VMs.
Data protection: Ensuring that data can be restored in the event of loss or corruption is a key IT requirement, made far easier by hyperconverged infrastructure.
The Seven Key Concepts Behind the hyperconverged Infrastructure
VM Mobility: Hyperconvergence enables greater application/workload mobility.
High availability: Hyperconvergence enables higher levels of availability than possible in legacy systems.
Data efficiency: Hyperconverged infrastructure reduces storage, bandwidth, and IOPS requirements.
Cost efficiency: Hyperconverged infrastructure brings to IT a sustainable step-‐based economic model that eliminates waste.
The driving force behind the implementation and adoption of hyper-‐converged infrastructures is largely two-‐fold. The advent of server virtualization – which has already been going on for the better part of a decade – and the more recent rise of flash memory as a storage media.
The Atlantis HyperScale CX-4 ApplianceLeverages main memory as a general-‐purpose primary storage tier that includes consistently high performance from any type of backing storage. Special algorithms identify transient I/O’s, preventing them from ever hitting other storage tiers reducing I/O traffic by up to 90% in some applications.
Hyperflex (Springpath)Combines the entire cluster’s SSDs and spinning disks into a single, distributed, multitier, object-‐based data store featuring real-‐time, always-‐on deduplication and compression to help cut storage costs
The Cisco HX Data Platform Controller
DataDirect Networks (DDN) SFA14KE
Leverages a combination of high-‐performance Intel processors, embedded networking, and NVMe SSDs to deliver 60 GBs per second of throughput and 6 million IOPS. That translates to 600 GB/PS and 60 million IOPS in a single rack.
EMC VCE VxRailVxRail Appliances bring together EMC's data services and systems management capabilities with VMware’s hyper-‐converged software -‐-‐including VMware vSphere, vCenter Server and VMware Virtual SAN
Gridstore HyperConverged Appliance
Gridstore includes compute and storage in a single system, and allow nodes to be either compute and storage or storage-‐only. The solution scales from three nodes to 256 nodes. The software architecture delivers native Windows integration, per-‐virtual machine I/O control, and elastic and independent scaling of resources.
Hitachi Data Systems Hyper Scale-Out Platform
Condenses computing, storage, networking and server virtualization into a hyper-‐converged scale-‐out appliance to implement a data lakes. Integrates high-‐performance NFS storage with Hadoop, Spark and Pentaho software to provide users insight from mixed data-‐in-‐place analytics workloads.
HPE Hyper-Converged Infrastructure
The Hewlett Packard Enterprise offering, the Hyper Converged 380, combines HPE's enterprise-‐class DL380 servers and a software stack that includes HPE's StoreVirtual VSA software-‐defined storage application and the HPE OneView converged management and automation application.
Nutanix AcropolisNutanix supports multiple hypervisors VMware, Hyper-‐V and KVM, including one-‐click conversion and cross-‐hypervisor disaster recovery and backup. Nutanix Prism, a purpose-‐built management tool, provides users with a "single pane of glass" for robust non-‐disruptive compute and storage operations.
Oracle SuperCluster M7The solution incorporates advanced processor-‐level security features, and works in a range of high-‐performance workload environments, from traditional ERP, CRM and data warehousing to e-‐commerce, mobile applications and real-‐time analytics. Based on Oracle's Sparc M7 processor, includes the Solaris operating system, Oracle VM Server for Sparc and the SuperCluster Management Toolset.
vSTAC OS natively converges storage, compute and virtualization resources (NexGen) into a modular building block that can be rapidly deployed to run any enterprise application.Solutions are either a preconfigured appliance or a software-‐only option that can be installed on the customer’s hardware platform of choice.
Pivot3 vSTAC
SimpliVity OmniStackStorage, compute, network efficiency, WAN optimization, data protection and performance with global unified management capabilities all packed into a 2U, x86-‐based server with sharable resources.In addition to SimpliVity's own OmniCube hyper-‐converged appliance, the company provides its software stack to technology partners like Cisco and Lenovo.
Scale Computing HC3Integrates storage within the kernel to streamline the I/O path for virtual machines while decreasing the required hardware resources. HyperCore intelligence automates the components which are clustered together into a single unified and redundant system to operate similar to a redundant and elastic private cloud
StarWind HyperConverged Appliance
StarWind Virtual SAN starts just on two nodes. It also possesses such advanced performance-‐related and cost-‐saving features as in-‐line deduplication to handle write-‐intensive workloads, Log-‐structured File System to eliminate small random writes typical for VM workloads, aggressive write-‐back RAM-‐based cache, inter-‐node tiering. Leverages Windows Server Storage Spaces Direct S2D
Stratoscale Symphony is a software-‐only hyper-‐converged infrastructure solution that runs on any commodity server hardware. The software is based on a microservices architecture. Microservices are distributed across the nodes of the cluster for resiliency, high availability and continuous scalability.
Stratoscale Symphony
Nutanix and Simplivity – have struck deals with server vendors such as Dell, Super Micro, Cisco, HPE and Lenovo, enabling them to offer a hyperconverged solution based around partners’ server hardware.
Partnerships within the HCI Marketplace
In Memory Processing
PernixData brings in IP and engineering talent relevant to the storage-‐class memory systems idea.
Pernix Data has two main products: one designed to accelerate storage performance (FVP) using server-‐side flash and RAM caching, the other to analyze storage performance across a cluster of VMs, servers and storage arrays (Architecture) to find performance bottlenecks.
Automation & Orchestration
The combination of Nutanix and Calm.io software allows customers to select the right private or public cloud for an application. Nutanix’ software stack will get cloud automation and management capabilities to “ deliver application and service orchestration, runtime lifecycle management, policy-‐based governance, comprehensive reporting and auditing services to support all application environments, including virtual machines, containers and micro-‐services.”
“Will this acquisition two-‐step, cause repercussions in the HCI vendor club as Nutanix’s competitors see a need to strengthen their own offerings? There is clear potential for this.”
29 Aug 2016
NVMe, the Insanely Fast Future for SSD’s
Like SCSI and SATA, NVMe is designed to take advantage of the unique properties of pipeline-‐rich, random access, memory-‐based storage. The spec also reflects improvements in methods to lower data latency since SATA and AHCI were introduced
M.2 SSD Drive Format
A prime example of the SSDs evolution are new M.2 form factor SSDs. The drives can help meet the needs of demanding enterprise applications including online transaction processing, high-‐performance computing and big data analytics. Data and time are money with ever-‐increasing amounts of data, M.2 provides the ability to quickly access and process it to continue deriving value
The 25 Gigabit Ethernet Consortium
Their Goal was to Develop specifications for 25 GbE and 50 GbE, designed in such a way that the group expects to make the cost of 25 GbE equal to 10 GbE.
Two words represent what makes 25 GbE a game-‐changer: single lane.
The consortium's specification for 50 GbE uses two lanes of 25 Gbps.
The second generation of 100Gbps is built on a totally new signaling scheme: four individual lanes of 25 Gbps. Reducing the number of pins on the chip makes it less expensive to produce and less power-‐hungry.
Are NVMe Fabrics in Your Future?
It's fair to say that RDMA over Ethernet is still in the early adopter stage, but to understand its future one has to look at some trends in the wider industry.
Memory Semantic Fabric
Gen-Z is a high performance, low latency, memory-semantic fabric that can be used to communicate to every device in the system. Gen-‐Z’s peer-‐to-‐peer architecture works with a diverse set of devices such as processors, memory, storage, FPGAs, , I/O, etc. This will allow any device to communicate with any other device as if it were communicating with its own local memory.
Speak the Cloud’s Language!
Offer the ability to provision new resources and control them via software APIs.
Cloud Bursting involves offloading compute and storage from on-‐premise architecture to public cloud infrastructure when your local infrastructure feels the strain.
Resources on Demand
Cloud Bursting is a way of using public cloud resources to deal with spikes in demand.
Cloud Bursting is a way of dealing with peaks in IT demand that can't be handled by on-‐premise hardware. This technique is useful to organizations with on-‐premise private clouds.
Ability to cope with variable computing demand by maintaining only enough hardware for predictable loads and shipping the rest of the workload — the spikes in demand — to the cloud
Automating the Convergence
You’re writing scripts to talk to manufacturer’s API’S and bring back that data and push it through another set of APIs to get it into a vendor’s tool. That management automation is a big part of the hyperconverged world
The likes of Puppet, Chef and SaltStack are all helping to bring these APIs together
Game Changing Ideas
Microsoft has that unique position in the marketplace of providing highly scalable and globally distributed public cloud services (in Azure public), while also providing organizations the ability to run the exact same applications, capabilities, and cloud scalable and redundant services on premise.
Who’s on First
With different teams handling the network, storage, server and security, which team should be assigned overall responsibility of the hyperconverged hardware.
With hyper convergence, there is no ownership of this piece of hardware; it is not a server, storage or network by itself, there is a need to establish an infrastructure team with the right combination of skills to adequately oversee the new environment.
Which Path to Success?
A truly effective infrastructure will differ from business to business. Organizations must thoroughly consider costs, time to production and compliance factors that will suit their specific requirements and lay the path to the best possible solution for their business.
Balancing the Innovations of HCI
A business must also be sure to assess and outline the acceptable level of risk with such an infrastructure overhaul. Is the organization open-‐minded to new ways of doing things or will it principally only ever work with tried, tested and established technologies?
The Value of Predictable Results
Pre-‐integrated, pre-‐built, pre-‐tested and pre-‐validated solutions provide predictable performance at a predictable cost-‐point with predictable outcomes. Therefore, this removes a lot of the guess-‐work typically associated with infrastructure projects.
Deploying Compatibility with Ease
With hyper-‐converged systems, deployment is much simpler. Rather than an organization having to do the hard work of determining how to piece systems together by studying all the compatibility matrices, the work has already been done by the experts, cutting time to production.
One Size Doesn’t Fit All
The biggest challenge for hyperconverged vendors is in the way they scale, by adding new nodes to the infrastructure. While this is part of the simplicity of the hyperconverged approach, adding an entirely new node can lead to some resources being underutilized if a workload simply requires more memory, for example.
Tailoring the Hardware
The capability of Independent scaling of the computing, caching, and capacity tiers, ultimately gives you the flexibility to scale the environment based on evolving business needs and resource consumption patterns
The Future Process of Evolution
HCIS use cases have so far been limited, causing silos with existing infrastructure, according to Gartner. Its progression will be dependent on multiple hardware and software advances, By 2019, approximately 30% of the global storage array capacity in enterprise data centers will be deployed on software-‐defined storage (SDS) or hyperconverged integrated system (HCIS) architectures.
Blurring the Lines of Distinction
Ultimately, the underlying infrastructure will disappear to become a malleable utility under the control of software intelligence and automated to enable IT as a service (ITaaS) to business, consumer, developer and enterprise operations.
The Finish Line, The Final Thought
Don’t simply choose a product. Meeting business requirements is the most important Objective. IT is just a Conduit to serve the Business Goals and to Provide that Competitive Advantage