I want to create a fairly simple mathematical model that describes usage patterns and system trade-offs.
The system behaves as follows:
- clients periodically issue multi-sheet packets to the host network.
- any host that receives a packet responds to a unicast response directly
- the originating node caches responses for a given period of time, then discards them
- if the cache is full the next time a request is required, the data is retrieved from the cache and not from the network.
- packets are fixed in size and always contain the same information
- hosts are symmetic - any host can issue a request and respond to requests
I want to create some simple mathematical models (and graphs) that describe the trade-offs available with some changes to the above system:
- What happens when you change the amount of time that the server caches responses? How much data does this save? How many calls to the network do you avoid? (clearly dependent on activity)
- Suppose answers are also multi-leafed, and any host that eavesdrops on another client request can cache all the answers it hears, thereby preserving itself, potentially making a network request — how will this affect the general state of the system?
- - - , , . , , "" ( , ) ? , .
, , . , , , , , , .
.