Subject: SD beginner : immaterial flow modeling for computer science topic ?

Posted by Martin Schaffernicht on 6/17/2010
In Reply To:SD beginner : immaterial flow modeling for computer science topic ? Posted by Michel Lutz on 6/17/2010

Message:

Hi again,

I guess my previous post came to you late because even though I sent it off last night, I had to confirm the post - which I discovered this morning. A real-life example of a delay...

OK, I see I had an awful mental model of a CPU! Even so: if what you have is not Bytes but Cycles, still there will be a certain number of Cycles "in use" or "assigned" and others "not in use" (or something like
that): each of the Cycles will be in one of the two possible states at any point in time. Since there are no other relevant differences between the Cycles, we can aggregate them into two groups or accumulations (or stocks), which may well have the respective state's name.

This is like a fixed population, and there are two transitions from one state to the other; these would be the flows, and they might have names like "assigned" end "freed" and would be "cycles per millisecond" or a similar short period of time. Thus you cannot assign more Cycles than you have in the computer, so your percentage will alwas be in between 0 and 1.

If you hace a certain disc space for the database, you may repicate this structure between "free" and "used".

Of cause, total disc space need not be eternally constant: you may design a rule that monitors a disc occupation indicator and triggers expansions. This could be done also if you divide the disc space into distinct portions, for instance one space for a database and a different space for other stuff. Then of cause you would have to consider 4 accumulations (free/occupied for database/other stuff) and certain flows might permit dynamic allocation of space to the database or the "other stuff".

Again, this is only how I would intuitively start representing it.

Best greetings,
Martin