[cs63201] on the operation of vector clocks in BSS

Mikhail Nesterenko mikhail at cs.kent.edu
Thu Mar 4 13:58:11 EST 2010


AOS students,

I looked again in the original paper by Birman, Sipher and
Stephenson. The paper is in the materials section. It appears that
they do not use the "true" vector clock increment. That is, when the
new value is selected, the local counter is not the maximum of all
values + 1. Instead, the local counter is just incremented. Hence, the
local counter is indeed the number of broadcasts each proces did.

The slides' description and the diagram I drew are correct.

Thanks,
--
Mikhail


More information about the cs63201 mailing list