Even so multiple data centers makes for a challenging

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: axos we can coordinate actions such that we are correct in all cases, and We then looked at a set of case studies: Facebook illustrated typical sharded architecture: front- ends assemble the web page, memcache provides good performance in the common case, and the storage server is the backstop. Even so, multiple data centers makes for a challenging design problem. We spent a while diving into Google’s architecture: Chubby – a lock server and cache manager based on Paxos MapReduce and Spark – a simple programming model for data- oriented apps GFS – large scale block storage with app- specific recovery BigTable – a key- value store, where related data is stored together, using log- deltas for efficient storage and lookup Spanner – extends BigTable to multiple data centers, using real- time for better performance And we add...
View Full Document

{[ snackBarMessage ]}

Ask a homework question - tutors are online