Hey everybody,
I decided to write about some trends in these first two papers that seem to be generally important concepts in distributed systems.
These were laid out plainly in the E. Brewer, Lessons from Giant-Scale Services while the M. Freedman, Experiences with CoralCDN: A Five-Year Operational View was more of a case study highlighting these similar aspects. A primary concern of these systems is reliability which takes form in fault tolerance. There is much discussion of how to deal with failing nodes and even disasters in these systems. These same issues are a focus of the CoralCDN reflection as they must deal with websites in flash-crowd scenarios that may not necessarily be able to respond to requests. Another obvious focus is the minimizing of client latencies or simply making things faster, a primary focus of distributed computing. As well, noted in both papers was minimizing internal bandwidth usage and utilizing resources as effectively as possible. This was discussed in the DQ principle as bandwidth correlates more than I/O seeks with overall performance capacity.
I decided to write about some trends in these first two papers that seem to be generally important concepts in distributed systems.
These were laid out plainly in the E. Brewer, Lessons from Giant-Scale Services while the M. Freedman, Experiences with CoralCDN: A Five-Year Operational View was more of a case study highlighting these similar aspects. A primary concern of these systems is reliability which takes form in fault tolerance. There is much discussion of how to deal with failing nodes and even disasters in these systems. These same issues are a focus of the CoralCDN reflection as they must deal with websites in flash-crowd scenarios that may not necessarily be able to respond to requests. Another obvious focus is the minimizing of client latencies or simply making things faster, a primary focus of distributed computing. As well, noted in both papers was minimizing internal bandwidth usage and utilizing resources as effectively as possible. This was discussed in the DQ principle as bandwidth correlates more than I/O seeks with overall performance capacity.
This comment has been removed by the author.
ReplyDeleteHi, I dont' have the privilege to add a new post, so I write here.
ReplyDeleteLessons from Giant-Scale Services is a paper published more than 10 years ago, but the arguments are still valid. For example, the basic model for giant-scale services nowadays still apply the model described in this paper: Clients connect via the Internet and then go through a load manager that hides down nodes and balances traffic. On the other hand, we have built more efficient and sophisticated mechanisms to replace the traditional "round-robin DNS". In Lessons from Giant-Scale Services, the author proposed a look-up model by introducing a "load manager" in layer 4. Now we have a new open content distribution network based around peer-to-peer technologies, called CoralCDN, introduced in Experiences with CoralCDN: A Five-Year Operational View. Coral’s three-level hierarchical overlay structure, a node would always try to query the nodes which have shorter distance, as a result, congestion in some nodes could be prevented. As we see, CoralCDN is a successful application of the idea of distributed systems.