P4L2 distributed file system Flashcards
Why it matters
achieve high performance in its file system, permits files to be shared by multiple clients without danger of stale data, client caching reduces server loading by 50% and network traffic by 75%.
The Sprite caching paper motivates its design based on empirical data about how users access and share files. Do you understand how the empirical data translated in specific design decisions? Do you understand what type of data structures were needed at the servers’ and at the clients’ side to support the operation of the Sprite system (i.e., what kind of information did they need to keep track of, what kids of fields did they need to include for their per-file/per-client/per-server data structures). (TO DO)
.
What are some of the design options in implementing a distributed service?
What are the tradeoffs associated with a stateless vs. stateful design?
What are the tradeoffs (benefits and costs) associated with using techniques such as caching, replication, partitioning, in the implementation of a distributed service (think distributed file service).
NFS semantics
Sprite semantics
See what problem they’re setting up and how they approach it.
Problem
- Caches have been used in many operating systems to improve file system performance.
how they approach it
- places at once. This is done through a simple cache consistency mechanism that flushes portions of caches and disables caching for files undergoing read-write sharing.
- The second unusual feature of the Sprite caches is that they vary in size dynami-cally. This was a consequence of our desire to provide very large client caches, perhaps occupying most of the clients’ memories.
conclusion. See what they found
- Sprite’s file system demonstrates the viability of large caches for providing high-performance access to shared file data. L
- . For users considering the purchase of a local disk, our advice is to spend the same amount of money on additional memory instead.
