Создание HPC кластеров с использованием IBM iDataPlex

Весьма познавательный RedBook посвящённый созданию высокопроизводительных кластеров на базе IBM iDataPlex. Уникальные характеристики этого решения достойны всестороннего изучения. Низкое энергопотребление. Малая занимаемая площадь. Небольшое тепловыделение. Высочайшая производительность. Всё это – IBM iDataPlex.




The IBM® iDataPlex™ data center solution is for Web 2.0, high performance computing (HPC) cluster, and corporate batch processing customers experiencing limitations of electrical power, cooling, physical space, or a combination of these. By providing a big picture approach to the design, iDataPlex solution uses innovative ways to integrate Intel® -based processing at the node, rack, and data center levels to maximize power and cooling efficiencies while providing necessary compute density.

An iDataPlex rack is built with industry-standard components to create flexible configurations of servers, chassis, and networking switches that integrate easily. Using technology for flexible node configurations, iDataPlex technology can configure customized solutions for applications to meet specific business needs for computing power, storage intensity, and the right I/O and networking.

This IBM Redbooks® publication is for customers who want to understand and implement the IBM iDataPlex solution. It introduces the iDataPlex solution and the innovations in its design, outlines its benefits, and positions it with IBM System x and BladeCenter® servers. The book provides details of iDataPlex components and the supported configurations. It describes application considerations for an iDataPlex solution and aspects of data center design that are central to that solution. The book concludes by introducing the services offerings available from IBM for planning and installing an iDataPlex solution.


Chapter 1. Introduction
1.1 Data center facts
1.1.1 Project Green IT
1.1.2 Next generation data center model
1.1.3 New philosophy
1.2 Introducing iDataPlex solution
1.2.1 The iDataPlex rack
1.2.2 Rear Door Heat eXchanger
1.2.3 FlexNode technology
1.2.4 Managing the iDataPlex systems
1.2.5 Onboard manageability
1.3 Support for GPGPU-based computing
1.4 Summary

Chapter 2. Benefits of iDataPlex
2.1 Acquisition and operating costs
2.1.1 Acquisition
2.1.2 Operating costs
2.2 Integration-ready
2.3 Modularity and flexibility
2.4 Rack density
2.5 Density at the data center level
2.6 Rear Door Heat eXchanger option

Chapter 3. Positioning
3.1 Data center-level solution
3.2 iDataPlex customers
3.2.1 Web 2.0 data center
3.2.2 High performance computing clusters
3.3 Hardware redundancy
3.4 Scale-up versus scale-out
3.5 Design objectives by platform
3.5.1 BladeCenter
3.5.2 System x
3.5.3 iDataPlex
3.6 IBM Intelligent Cluster and iDataPlex
3.7 Positioning iDataPlex solution with BladeCenter

Chapter 4. iDataPlex products
4.1 The big picture of iDataPlex
4.2 iDataPlex rack components
4.2.1 The iDataPlex rack
4.2.2 iDataPlex cable brackets
4.2.3 iDataPlex cooling options
4.2.4 Rear Door Heat eXchanger
4.3 iDataPlex FlexNode chassis components
4.3.1 2U and 3U chassis
4.3.2 Direct Dock Power
4.3.3 Chassis power supply, power cable, and fans
4.4 iDataPlex servers and components
4.4.1 Comparing the iDataPlex servers
4.4.2 Processor options on the iDataPlex servers
4.4.3 Memory options on the iDataPlex servers
4.4.4 iDataPlex dx360 M2 server
4.4.5 iDataPlex dx360 M3 server
4.4.6 Operator panel
4.4.7 Chassis trays
4.5 Storage and RAID options
4.5.1 Internal storage options
4.5.2 Solid-state storage
4.6 NVIDIA graphics adapter and GPGPUs
4.6.1 Quadro FX 3800
4.6.2 Tesla M1060 and Tesla M2050
4.7 PCI Express I/O adapters
4.8 Ethernet switch options
4.9 InfiniBand options
4.9.1 InfiniBand speed terminology
4.9.2 InfiniBand connectors and cables
4.9.3 InfiniBand versus 10 Gbps Ethernet
4.9.4 QDR HCAs and switches
4.9.5 QLogic InfiniBand Fabric Suite 6.0
4.10 PDU options
4.11 iDataPlex chassis configurations
4.11.1 Compute-intensive server (1A and 1B)
4.11.2 Storage server (2A, 2B, and 2D)
4.11.3 I/O-intensive server (3A, 3B, and 3D)
4.11.4 3U chassis configurations (4A)
4.11.5 General hardware configuration rules
4.12 iDataPlex Management components
4.12.1 IPMI and IMM
4.12.2 UpdateXpress System Packs
4.12.3 IBM Systems Director Update Manager
4.12.4 Other firmware update methods
4.12.5 Update notification subscription

Chapter 5. Application considerations
5.1 Hardware constraints
5.1.1 Chassis considerations
5.1.2 iDataPlex server considerations
5.1.3 Stateful versus stateless operation of the nodes
5.1.4 Conclusions for operating systems and applications
5.2 Operating system considerations
5.2.1 Supported operating systems
5.2.2 Location of the operating system
5.3 Application considerations
5.4 Redundancy considerations
5.5 Example iDataPlex scenarios
5.5.1 Web 2.0 example
5.5.2 HPC example

Chapter 6. Designing your data center to include iDataPlex
6.1 Data center design history
6.2 Rack specifications
6.3 Rear Door Heat eXchanger specifications
6.3.1 Planning for the arrival
6.4 Cabling the rack
6.4.1 Nodes
6.4.2 Management appliance
6.4.3 Switches
6.4.4 Power distribution units
6.4.5 Cabling best practices
6.4.6 iDataPlex overhead cable pathways
6.5 Data center power
6.5.1 Aggregation and reduction of data center power feeds
6.6 Cooling
6.7 Airflow
6.7.1 Heat pattern analysis without the Rear Door Heat eXchanger
6.7.2 IBM Systems Director Active Energy Manager
6.7.3 Hot air ducts
6.7.4 iDataPlex and standard 1U server rack placement
6.7.5 Heat pattern analysis with the Rear Door Heat eXchanger
6.7.6 Rear Door Heat eXchanger airflow characteristics
6.8 Water cooling and plumbing
6.8.1 Technical specifications
6.8.2 Absorption rates
6.9 Designing a heterogeneous data center
6.10 iDataPlex servers in a standard enterprise rack
6.11 The Portable Modular Data Center
6.12 Possible problems in older data centers
6.12.1 Power
6.12.2 Cooling
6.12.3 Static electricity
6.12.4 RF interference
6.13 Links to data center design information

Chapter 7. Managing iDataPlex
7.1 Integrated Management Module
7.2 Unified Extensive Firmware Interface
7.3 Dynamic System Analysis
7.4 Advances Settings Utility
7.5 SMBridge
7.6 The Avocent rack management appliance
7.7 IBM power distribution units
7.8 IBM Systems Director
7.9 xCAT
7.10 Distributed Image Management
7.11 Cluster resource management
7.11.1 Moab Adaptive HPC Suite
7.11.2 Moab Cluster Suite

Chapter 8. Services offerings
8.1 Global Technology Services
8.1.1 High-density computing data center readiness assessment
8.1.2 Thermal analysis for high-density computing
8.1.3 Scalable modular data center
8.1.4 Data center global consolidation and relocation enablement
8.1.5 Data center energy efficiency assessment
8.1.6 Optimized airflow assessment for cabling
8.1.7 IBM Server Optimization and Integration Services: VMware
8.1.8 Portable Modular Data Center
8.2 IBM Systems and Technology Group Lab Services
8.2.1 Data center power and cooling planning for iDataPlex
8.2.2 iDataPlex post-implementation jumpstart
8.2.3 iDataPlex management jumpstart
8.2.4 Cluster enablement team services
8.3 IBM Global Financing


Оставьте комментарий