[dba-Tech] A Fast Wait-Free Hash Table

Jim Lawrence accessd at shaw.ca
Mon May 10 00:11:20 CDT 2010


How do these super fast databases work? Super databases like 'Cassandra'.

Through a single system with a single CPU the process is fairly straight
forward but now add multiple CPUs and to that the potential of multiple
computers. The question is can multi-threaded applications be designed, that
can grow in real-time, potentially have unlimited data and still perform at
super speeds?

The whole operation hinges on the ability of a system to generate unique
hash key pairs at incredible speed and the subsequent ability to retrieve
information via those keys. The following link is to a series of YouTube
lectures given at Stanford University, some by Cliff Click, a senior
developer for the Azul project (http://www.ccpalma.com). He gives the coding
examples in Java but they are still easy to follow as the code is kept
simple and VB/C like.

In conclusion this links to a great series lectures:
http://www.youtube.com/watch?v=WYXgtXWejRM

Jim 




More information about the dba-Tech mailing list