Business Plan Overview Grid Dynamics

Reviews
Shared by: Dudi Einey
Stats
views:
14
rating:
not rated
reviews:
0
posted:
6/2/2009
language:
English
pages:
0
Project Convergence: Integrating Data Grids and Compute Grids Eugene Steinberg, CTO Grid Dynamics May, 2008 Data-Driven Scalability Challenges in HPC  Data is far away  Latency of remote connection  Latency of data movement through pipes  Chatty algorithms are expensive  Data is centralized     HW Resources are limited Inevitable disk I/O due to limited RAM Connections are limited Highly concurrent access doesn't scale well Grid Dynamics 1 Usual Solution: Compute Grid + Data Grid  Classic Data Grid      Data is partitioned Partitions are stored in memory of data grid Data grid is deployed near to compute grid Search is parallelized over partitions Build-in replication, persistence, coherence, failover  What is Achieved?     Reduced latency and data moving cost Improved connection scalability Reduced data contention No Disk I/O – 100% memory speed Is this the best we can do? Grid Dynamics 2 Limitations of Compute Grid + Data Grid  Two separate grid environments  Hardware, footprint and management costs of dual infrastructure  Segregated infrastructures cannot share resources Compute Grid  Sub-optimal resource utilization  Compute grid is CPU-bound, not RAM-bound  Data grid is RAM-bound, not CPU-bound Data Grid  Still sub-optimal performance  Still paying for remote network calls and data movement Grid Dynamics 3 Better Answer: “Compute-Data Grid”  Shared hardware between compute & data grid  Data grid resides in RAM of host machines  Compute grid runs HPC jobs on the same host machines  Opportunity to collocate processing with data  Many applications support compute-data affinity  No network overhead on remote calls and data movement  New recipe for scalability  As HPC application needs to scale in and out, data partitions are spread over larger or smaller pool of hosts Grid Dynamics 4 Project Convergence  Open source reference architecture for Compute-Data Grid  Goals  Pluggable architecture to support adapters for many grid products  Non-intrusive compute-data grid coordination  Library of adapters for popular commercial and open source grids  Key Use Cases  Data-aware job scheduling  Dynamic data grid right-scaling Grid Dynamics 5 Logical Architecture Scheduler Compute Grid Wrapper Client HOST HOST HOST Monitor Data Grid Convergence Grid Dynamics 6 Implementation  Core Components  Data Grid Monitor: service responsible for knowing Data Grid’s topology and state  Data Aware Wrapper: client side library which extends Compute Grid’s scheduling API to support data-aware job scheduling  Main Workflow     Client code submits the job using Data Aware Wrapper Data Aware Wrapper consults Data Grid Monitor Data Grid Monitor returns a set of hosts that are nearest to the data Wrapper submits the job to the Scheduler, requesting specific hosts  Variation on Configuration  Monitor can be a network service or embedded as a library Grid Dynamics 7 Demo – “Hello, World” Trading Analytics  Setup     4 DataSynapse Engines, 2 per host 2 GigaSpaces partitions Scheduler: DataSynapse GridServer Client app + embedded Monitor + Wrapper Engine Engine P1 GridServer Scheduler P2  Test data  Stores100,000 trades for 10 stock tickers  Partitioned by ticker  Job  Computes simple statistics about trades  A Job spawns 10 tasks, one task per ticker Wrapper Client Monitor  Task Scheduling Control Functions  Data-aware, random, or anti-data-aware Grid Dynamics 8 Demo Screenshot Grid Dynamics 9 Current Project State      Hosted by OpenSpaces.org Licensed under Apache 2.0 Latest version 0.1.1 (Apr 2008 release) Use case supported: data-aware job scheduling Available plug-ins:  Compute Grids: Data Synapse GridServer 5.0  Data Grid: GigaSpaces XAP Grid Dynamics 10 Project Roadmap  Support Additional Adapters  Convergence 0.2: GridGain (under development)  Convergence 0.3: Oracle Coherence  Convergence 0.4: Sun Grid Engine  Support Additional Use Cases  Dynamic data grid right-sizing  Call for Action  Please, join the project to help test and extend the system, or provide additional adapters http://www.openspaces.org/display/CVG/Convergence Grid Dynamics 11 Q&A Grid Dynamics 12 Thank You! Eugene Steinberg, CTO esteinberg@griddynamics.com

Related docs
Business Plan Overview Grid Dynamics
Views: 64  |  Downloads: 5
Dynamics
Views: 6  |  Downloads: 0
Grid Economics and Business Models
Views: 4  |  Downloads: 0
Team Dynamics and Conflict Resolution
Views: 702  |  Downloads: 88
1What is a Grid
Views: 104  |  Downloads: 7
Microsoft Dynamics Quick Start
Views: 73  |  Downloads: 9
A. BUSINESS PLAN OVERVIEW
Views: 8  |  Downloads: 0
MS Dynamics Nav - Solution Overview
Views: 7  |  Downloads: 0
Buyers_Guide_Grid
Views: 37  |  Downloads: 5
INTRODUCTION TO TEAMWORK AND GROUP DYNAMICS
Views: 32  |  Downloads: 1
premium docs
Other docs by Dudi Einey
Sesame Street - 40th Anniversary
Views: 49  |  Downloads: 0
100 Outstanding Marketing Tips
Views: 65  |  Downloads: 2
Your Guide to Good Health Insurance
Views: 121  |  Downloads: 0
Timeless Sales Strategies
Views: 142  |  Downloads: 0
Selling Your First Million
Views: 156  |  Downloads: 0
ACORN Investigation Jarramt Arrest Warrant
Views: 33  |  Downloads: 1
United Nations Nations Unies
Views: 164  |  Downloads: 0
UNITED £% NATIONS O Security Council
Views: 118  |  Downloads: 0
Security Council
Views: 158  |  Downloads: 0
A S
Views: 109  |  Downloads: 0
S
Views: 107  |  Downloads: 0