Peano Trees, Data Striping, and Distributed Computing Otto Borchert North Dakota State University
Overview Background Prototype Advantages and Disadvantages Demo! Peano Trees Data Striping Distributed Computing Prototype Advantages and Disadvantages Demo!
Peano Trees P1: | | | | 1 0 0 0 |||| 1101 NZ: 1 | | | | 1 1 0 0 |||| | | | | 1 0 0 0 |||| 1101 NZ: 1 | | | | 1 1 0 0 |||| 1101 Quad P1 NZ [] 1000 1100 [01] 1101 1101
Data Striping RAID Level 0 “Striping Units”
Distributed Computing Shared Memory Many Processors, One Memory Area Low Maintenance, Low Scalability Shared Disk Middle Ground Shared Nothing Many Processors, nothing shared High Maintenance, High Scalability
Prototype System Server Clients Waits for connections Divides and distributes work Receives new data and redistributes Clients Connects to server Writes data to hard drive (striping) Sends more information to server
Advantages Easy to read Easy to debug Relatively easy to modify Excellent Vector Implementation
Disadvantages Poor P-tree implementation Uses characters instead of bit streams No user interface Doesn’t convert tabular data to raster format
Demo!