High Performance Cloud

Network

F5 VIPRION platform

Maximize Large-Scale Application and Firewall Performance

With its industry-leading layer 4/7 throughput, connection processing, and SSL/ECCperformance, VIPRION efficiently manages the most demanding applications, offloads servers,and consolidates your Application Delivery Network. In addition, as an ICSA Labs Certifiedfirewall solution, F5 BIG-IP® Advanced Firewall Manager™ (AFM) on VIPRION provides native,high-performance network firewall services to protect public-facing websites and data centerapplications from distributed, multi-layer cyber attacks.

VIPRION high-performance and distributed denial-of-service (DDoS) protection capabilitiesare enabled through field-programmable gate array (FPGA) technology tightly integratedwith the F5 TMOS® technology and software.

F5 embedded Packet Velocity® Acceleration (ePVA) FPGA delivers:

  • High-performance interconnection between Ethernet ports and processors.
  • L4 offload, enabling leading throughput rates and reduced loads on software.
  • Hardware-accelerated SYN flood protection.
  • 6Hardware detection and mitigation of more than 100 types of denial-of-service (DoS) and DDoS attacks.
  • Support for F5 IP Intelligence Services, with blacklist, whitelist, and graylist capabilities.
  • Native network overlay (VXLAN/NVGRE) support.
  • Hardware-enabled DNS caching, which hyperscales responses for fast service and app delivery (B2250).
  • User selectable hardware profiles that enable different performance levels for targeted workloads. Initial profile options include optimized L4 throughput on select platforms for CGNAT or L4-centric traffic management solutions.

Compute

HPE Chassis & Blades

The HPE BladeSystem enclosure provides the entire cooling, power, and I/O infrastructure needed to support interconnect, modular server, and storage components. The HPE ProLiant server provides the flexibility to optimize core IT applications with right-sized storage for the right workload for a lower TCO.

Increased Power

The HPE BladeSystem enclosure is created for larger data centers with more dynamic data center environments and mid-sized data centers. By increasing power supply wattage, the HPE BladeSystem enclosure has enhanced the foundation for converged infrastructure.

Efficient Performance

The BladeSystem has Platinum Power Supplies, Intelligent Power Module, and Location Discovery Services that increase power supply wattage and mid-plane bandwidth for demanding workloads.

High Performance

The HPE server delivers up to 70% performance increase from other generations. Delivering flexibility to optimize core IT applications to lower TCO, this blade is designed for a wide range of configuration and deployment options.

Versatile

For increased deployment flexibility, the HPE server delivers embedded storage controller options such as HPE Smart Array P244br, HPE H244br Smart Host Bus Adapter, or the HPE Dynamic Smart Array B140i Controller.

Storage

Bluekiri Enterprise Cloud Storage delivers high-performance, low-latency disk support for virtual machines (VMs) with input/output (I/O)-intensive workloads. VM disks that use EC Storage store data on Hitachi flash arrays.

With this EC Storage, Bluekiri offers the ability to truly lift-and-shift demanding enterprise applications.

You can run performance-intensive database workloads in applications like SQL Server, Oracle, MongoDB, MySQL, and Redis, which require consistent high performance and low latency.

Hitachi Accelerated Flash

Intelligent Flash Performance and Data Center Efficiency

Delivering on enterprise demands for real-time customer engagement requires more than a fast storage array. It requires an end-to-end approach to data management that leverages both the storage operating system and flash media, to deliver low latency performance even as data levels grow exponentially.

Current all-flash arrays (AFAs) rely on performance management to be handled in the array controller along with all other operations, such as data reduction. As data levels increase, they can cause controller bottlenecks, sporadic response times and a poor customer experience. To offset this issue, IT organizations have been forced to make tradeoffs in the number of workloads or amount of data they store on an individual AFA. These decisions result in the need to deploy and support more systems, raising costs and increasing management complexity

Hitachi Accelerated Flash fundamentally changes this paradigm. By combining a flash-aware I/O stack within Hitachi Storage Virtualization Operating System (SVOS) with industry-leading intelligent flash modules (FMDs), HDS enables IT teams to consolidate more workloads and data onto a single system. At the same time, they can provide scalable high performance and sub-millisecond response times. And with no-penalty, inline compression in each FMD, performance gains are achieved at the lowest effective cost1 , supporting greater operational efficiency.

HAF benefits start with SVOS, which includes an enhanced, flash-aware I/O stack designed to improve low latency throughput, even as data scales to multiples of petabytes. Using patented express I/O and flash-specific data reduction algorithms, SVOS accelerates read and write operations for a lower response time. SVOS also includes FMD integration, enabling the offload of metadata management tasks from SVOS to FMD memory for scalable performance. In configurations with a large number of FMDs, SVOS has the ability to scale processing many times further than common AFAs can and deliver more predictable long-term performance.

FMDs are intelligent flash devices engineered to help improve performance and efficiency by actively working with SVOS. The second and third generation of flash modules (FMD DC2, FMD HD) are built from the ground up to support concurrent, large I/O enterprise workloads and enable hyperscale efficiencies. At their core is an advanced embedded multicore flash controller that increases the performance of multilayer cell (MLC) flash. They intelligently manage I/O and have the ability to prioritize application I/O requests over background tasks, such as garbage collection, to minimize potential latency. Each FMD also includes an inline compression offload engine and enhanced flash translation layer that delivers up to 80% data reduction (typically 2:1) at 10 times the speed of competitors. Because inline compression can be distributed and managed across FMDs, it gives HDS the ability to easily provide five times, even 10 times the data reduction processing capabilities of traditional AFAs for a better long-term performance profile.

Because of this, Hitachi Data Systems guarantees at least a two times savings in flash capacity, but you’re likely to get a lot more.

For virtual server environments, HAF can provide additional advantages beyond long-term scalable performance and efficiency. Leveraging SVOS quality of service (QoS) functionality and integration with VVols, HAF can be used to deliver more predictable customer experiences. SVOS QoS and VVols work together to allocate specific resources to a virtual machine (VM), including FMDs. This ensures that an application running on a VVol not only gets a specific level of performance, but also that it has dedicated FMDs for managing compression. By dedicating data reduction resources to an application, you deliver a more predictable level of latency.

For resiliency, the integrated HAF architecture provides full visibility to every NAND, including its current state and wear level. Using flash wear management, data rebalancing and data reduction to minimize writes, HAF helps extend wear limits.

Business Benefits

Leading Performance, Lowest Bit Cost, Highest Capacity and Extended Endurance

Hitachi built this new solid-state storage offering with a new 2U flash tray that holds up to 12 high-density flash module drives and associated interconnect cables (see Table 1). FMDs deliver outstanding value compared to enterprise solid-state disk (SSD). When compared to a small-formfactor 1.92TB SSD, it delivers:

  • Up to three times the per-device random read 8KB I/O performance.
  • Up to five times the per-device random write 8KB I/O performance.
  • Up to 70% lower effective1 bit price.
  • 60% lower response time at peak load.
  • 60 times faster formatting speed.
  • Superior data integrity.