Scaling architectures is invariably a discussion about performance. And yet scaling can be applied to any of the systemic qualities of an architecture. More importantly, it must be applied to all systemic qualities or you run the risk of scaling beyond one or more of their reasonable limits. This presentation will discuss the systemic qualities of an architecture and how they interact. It will also provide recommendations on how to measure and manage scaling systemic qualities as a system rather than individually.
Latency. The time it takes packets to flow from one part of the world to another. Everyone knows it exists. The second fallacy of distributed computing is “Latency is zero”. Yet so many designs attempt to work around latency instead of embracing it. This is unfortunate and in fact doesn’t work for large-scale systems. This presentation will address how latency impacts a distributed system architecture and provide techniques for making the architecture more resilient to latency.
This was an unscheduled, ad hoc presentation by Dan on the technical issues that arise from the nature of eBay, including its enormous database, its huge number of users and hits, and the need for careful timing on placement of and closure of bids. Dan described the architectural solutions to these problems, in what turned out to be the single most popular session of the entire conference.
© 1997 2008 Kovsky Conference Productions Inc.; All rights reserved.