Questions for 6.S897 lecture 4 (Storage 1, 09/22).
Email your answers to 6.s897staff@gmail.com.

1) Why is the record append operation in GFS only guaranteed to write data "at least once", as opposed to "exactly once"? In what situations would a record be written twice?

2) What are the tradeoffs in setting the tract size in FDS? That is, what are the pros and cons of making it larger vs smaller?

3) Are there any advantages to Google's single-master design over FDS?