Review of Scale and Performance

From: David Coleman (dcoleman_at_cs.washington.edu)
Date: Mon Feb 23 2004 - 16:25:56 PST

  • Next message: Ankur Rawat \(Excell Data Corporation\): "Andrew Distributed File system"

    Scale and Performance in a Distributed File System describes the experiences and improvements in the Andrew File System developed at Carnegie Mellon University.  When I read the first portion of the paper I kept thinking that the approaches chosen would not work well either for scaling or even low usage performance.  I was wondering why we were even reading this paper except possibly as an example of a bad design.  I then had to laugh at myself when I realized the point of the paper was actually to describe the successful improvements made to achieve the scalability and performance goals.

     The cache coherence strategy used is not nearly as strong as is used for memory cache coherence.  However, I’m not sure that it needs to be.  I’m a little uncomfortable with the burden of synchronizing file access being placed on the applications given one of the goals of the system is transparency.  I did like the transaction-style strategy for broadcasting updates.  The changes are only visible when the file is closed.  I also liked the choice of allowing existing open handles to continue using the older data.  I know of transaction support going into a new OS and that implementation will make the modifications visible to existing handles as soon as the transaction is finalized.  I believe that strategy to be flawed.  I prefer either invalidating the handles or using old data as was implemented.

     


  • Next message: Ankur Rawat \(Excell Data Corporation\): "Andrew Distributed File system"

    This archive was generated by hypermail 2.1.6 : Mon Feb 23 2004 - 16:21:45 PST