2

From Google File System paper, Application communicates with GFS Client by sending (filename, byteRange) and then GFS Client converts that byteRange to ChunckIndex. From the document it is unclear to me

  1. Does byteRange represent begin and end of bytes to read/write OR does it represent number of bytes to read/write?
  2. How does GFS Client convert byteRage to ChunkIndex? What is the formula behind it and what other variable(s) beside byteRange involved in this calculation?

Thank you,

SlyBits
  • 31
  • 2

1 Answers1

0
  1. byte range vs byte offset: the client translates byte offset NOT byte range to chunk index. byte range is used later when client sends read request to chunk server.
  2. translate byte offset to chunk index is simple: because chunk size is fixed at 64MB. say the client wants byte offset at 65MB, then the chunk index would be 1. and if client wants byte offset at 130MB, then the chunk index would be 2.
user2001850
  • 477
  • 4
  • 18