IBM PureSystems

Editor's Notes

• #3: IDC has been tracking IT spending for almost two decades, and a definitive trend has emerged. When we look back at 1996, this was the day of Big Iron. Servers were large and expensive. The initial purchase was hefty, but the management and administration costs were low. That’s because you don’t have that many of these monsters in your environment.Then, in 2001, distributed computing on x86 systems really took off. Servers cost dramatically less, but you had so many of them. There was a shift in IT budgets, where now you spent the majority of your dollars on maintaining this growing infrastrcuture. In 2006, we reached a tipping point, and very little was spent on new systems, but powering these massive farms, and manually integrating these systems into your environment was a huge burden. And so arose virtualization in the x86 space. By 2011, less and less was being spent on hardware, and virtualizing assets reduced the power and cooling costs somewhat, but administration of the environment…whether physical or virtual, was still consuming the lion’s share of IT budgets. And so that’s where we are today. But if you’re spending all your time and money on maintaining environments, how do you then find the time and money to deliver new technologies required to drive the business?That’s where integrated systems come in. <Button Click>. Integrated systems are designed to recapture the portion of your IT budget related to keeping the lights on. The goal is to repurpose that spend towards new technologies designed to drive revenue and value for your company. They’re designed to properly align your IT efforts with the needs of the business.
• #4: This is a supply and demand problem. The demand for people is growing. Customers now spend 3x to 4x more on people than server hardware. According to Gartner, the supply of qualified people due to retirements is reducing. These numbers are all industries but IT will be feeling this impact as it gets harder to find qualified people.
• #11: First, we'll start with the integrated infrastructure. IBM offers you one system that brings your resources together. It's designed for future generations of technology. It delivers energy efficient power and cooling, which is key to getting the maximum performance out of all the components over future generations.And, unlike a lot of other solutions, we support multiple architectures, but don't require a lot of expensive, non standard components. IBM PureFlex System provides incredible capacity for compute, memory, and storage with a high-level of IO operations in a single rack – with the ability to have up to four full racks in a single system.
• #15: No compromise designs for integrated compute nodes that support POWER7 or x86 architectures provide the performance behind the IBM PureFlex System.It allows you to share your network, storage and other resources across Power and X86 applications. And we've built in the security so that when you start the system up, it's doing all the validation and checks across the physical and virtual environments to reduce risk and save administrator time in configuring security.
• #29: http://www-03.ibm.com/systems/info/x86servers/serverproven/compat/us/serverproven/indexf.html
• #43: SW Integration for System x includes preloading VMWare ESXi hypervisor on USB key.
• #60: Topology patterns: These patterns are used to quickly and easily deploy database clusters. A cluster instance is a set of database nodes that work together to provide higher levels of scalability, availability and throughput. Using topology pat- terns, administrators can create database clusters in minutes, eliminating the time, effort and expertise required to do so manually. These patterns offer flexible options to create clusters of different sizes to best fit the application’s needs. Database patterns: These patterns are used to quickly and easily deploy DB2 databases on the system. The system includes a default OLTP transactional database pattern that can be used to create databases that are pre-tuned and pre- configured specifically for high-performance transaction processing. IT departments can also create their own database patterns by cloning existing databases. In both cases, database patterns enable quick, simple and consistent database deployment as well as provide a user-friendly, self-service environment for application development. Patterns help create a standardized infrastructure, reducing the risk of errors by encouraging optimized database configurations, reducing IT operational overhead and improving productivity for application developers. Built-in database clustering Software pre-installed in PureData System for Transactions provides automatic deployment, management and monitoring of database clusters, which supports fast, seamless failover. The system also delivers high scalability by enabling administrators to add more nodes with no application changes required. Plus, no database clustering skills are required to manage the system. Automatic workload balancing across the multiple servers in a cluster boosts throughput for optimum performance, and higher utilization is provided through shared resource management to help lower costs. Built-in, automated storage tiering The virtualized storage in PureData System for Transactions is designed for fast performance and to rapidly respondto business needs, being powered by IBM Easy Tier® technology. The system automatically manages hot, warm and cold data to enhance performance and cost savings without sacrificing data accessibility. Topology Patterns. Topology patterns will install all of the software required to run pureScale and create the pureScale instance on a number of compute nodes. You can deploy a topology pattern of 2, 4, or 6 nodes. The 2 node topology pattern consists of an instance of 2 members with a cluster caching facilities (CFs) co-located on each of two compute nodes. This is the smallest practical HA installation of Db2. The 4 nodes topology pattern consists of 2 members and 2 CFs on Ratings 0 separate compute nodes. The 6 node topology pattern is a 4 member cluster with 2 CFs on separate compute nodes. Of course the more nodes you deploy to the higher the performance and resilience.Database Patterns. A database pattern is essentially a method of storing and reusing database configuration settings. Databases patterns are used to create and configure a database within the instance topology. PureData systems will come with an IBM transaction processing database pattern. You can also "roll your own".
• #64: This is an overview of the PureData System for Analytics. While you can but PureData for Analytics in ¼ rack, ½ rack, full rack, and multi-rack configurations, each rack will look like this. There are 8 SAS disk enclosures with a total of 96 1TB drives, two front end “hosts” using dual quad core processors and seven 146Gb drives each. But the real work happens on the S-Blades. In a quarter rack configuration there are 4 S-Blades, in a half rack there are 7, and in a full rack there are 14. These S-Blades have two quad core CPUs and four dual core Field programmable gate arrays (FPGAs) along with 24GB of memory. The FPGAs act as an intelligent query filter -- automatically discarding the 90-95% of the data that is irrelevant to the query being run. The built in expertise of the The FPGA does more than that though… certain types of processing are pushed down to run inside the FPGA -- on the fly decompression, projections, restrictions, and visibility lists are all handled within the FPGA. This patented hardware acceleration layer is the secret sauce that provides the breathtaking performance expected from IBM PureSystems. All of this adds up to a system that can handle up to 128GB of data in a full rack (Over a petabyte in a 10-rack system) that can scan data at 145TB per hour and load data at over 5TB per hour --- All without any need for indexes, storage management, or tuning.
• #101: Remove power connector