Welcome in Dinesh Patel Blog

Why Analytics? BI Reporting and Ad-Hoc Analytic What happen? When and Where it happened? How much? Predictive Analytics What will happen? Where will the impact be? Optimization  What is the best choice? Challenges of Analytics Expensive Inefficient Process Limited analytic complexity Inability to react to market conditions Time consuming Inability to experiment Processing on limited and stale data

Netezza: ======== Netezza is rows base database using colummar compression Hash algorithm is used S-Blade / Snippet Processing Unit (SPU): ======================================= Blade Server Card and Netezza DB Accelerator (FPGA) 1 Blade Server Card = 8 Core CPU (2 Intel Quad-Core) 1 Blade Server Card = 16 GB DRAM 1 Netezza DB Accelerator = 8 Core FPGA (4x Dual-Core) 1 Snipet Processing Unit = 8 Disks (responsible for) 1 Snipet Processing Unit = 8 Snippet Processor 1 Snippet Processer = 1 Core CPU, 1 Core FPGA and 2 GB DRAM 1 S-Blade = 6 sisks and 2 spare disks Snippet Processing Array (SPA): =============================== 1 Storage Processing Array (SPA) = 4 Disk Encloures 1 Storage Processing Array (SPA) = 2 Spares Disks 1 Disk Encloures = 12 Disks 1 Disk = 1 TB SAS Drive 1 Disk = data on drive call data slice (table are splits across data slides 1 Disk = Devided into three partion - Primary, Mirror and Temp Failed drive is automatically regenerated on a ...

http://www.pentaho.com/Optimize-the-Data-Warehouse