 |
 |
Hi Pete, > Other than the current implementation imposes why did you chose the limit > of 64 servers? (i.e. did you do any quantitative estimate that gave that > as the order of magnitude at which partitioning the system made sense?) The cost of handlinbg more that 64 servers increases the cost substantially but in terms of server selection and memory to keep the extra information around. It just didn't make sense. 64 looks like a good number as well as fitting into a good data type and a good scaling turn-over point. > What else scales with the number of server olbd's? I don't understand the question. Andy