CloudComputing - coursework + projects
Cloud-systems coursework spanning MapReduce, distributed key-value stores, and graph-scale workloads. Includes a from-scratch MapReduce coordinator + worker with plugin-loaded map/reduce functions, distributed kv-store experiments, and a generated mini-internet used to benchmark PageRank locally vs. on Google Cloud. Repo serves as the working scratchpad for BU cloud / distributed-systems labs.
Highlights
- MapReduce coordinator + worker in Go with plugin-loaded map/reduce functions
- Distributed key-value store experiments (replication, leader assignment)
- Generated-HTML mini-internet for PageRank benchmarking
- PageRank evaluated locally vs. on GCP — same code, two scales
Tech
The canonical source for this project is on GitHub.
View on GitHubREADME · github.com/ArkashJ/CloudComputing
CLOUD COMPUTING
Most of today's organizations needing a technology solution look to satisfy their computing, storage and networking needs through one of the large public cloud providers. Unlike traditional environments where a company had to build its own infrastructure often at large time and monetary expense it can now rent what it needs at the click of a button. In this course we will provide hands on experience with one of the large public cloud platforms. In particular we will look into the different flavors of compute, storage and networking available, how best to use them to solve interesting problems, and how to do everything on a constrained budget. Students will get accounts and deliver project work on the public cloud while also learning some of the fundamental principles on how those different cloud systems work under the covers