Generational garbage collection and the radioactive decay model

Abstract

If a fixed exponentially decreasing probability distribution function is used to model every object's lifetime, then the age of an object gives no information about its future life expectancy. This radioactive decay model implies there can be no rational basis for deciding which live objects should be promoted to another generation. Yet there remains a rational basis for deciding how many objects to promote, when to collect garbage, and which generations to collect. Analysis of the model leads to a new kind of generational garbage collector whose effectiveness does not depend upon heuristics that predict which objects will live longer than others.This result provides insight into the computational advantages of generational garbage collection, with implications for the management of objects whose life expectancies are difficult t,o predict.

Allocation-phase aware thread scheduling policies to improve garbage collection performance

“…We use mark/cons ratio [4,7,11] to measure the GC overhead. Mark/cons ratio is defined as the total number of bytes copied during all garbage collections divided by the total number of allocated bytes.…”

Using page residency to balance tradeoffs in tracing garbage collection

“…An alternative heuristic would measure age not in terms of the number of collections survived but in terms of bytes subsequently allocated. Following the intuition of older-first garbage collection [13] our collector could predict that the residency of pages allocated just before a collection will be higher than those allocated earlier, possibly promoting the former in place (depending, of course, on the evacuation threshold).…”

On models for object lifetime distributions

“…In a garbage collector this property assures that an object just discovered live by the collector has the same residual lifetime as the lifetime of a newly allocated object, and this greatly simplifies the analysis. Thus, the exponential distribution was used by Clinger and Hansen in the analysis (and to inspire the design) of a non-predictive collector, outside the generational realm [3]. In our examination of a generalized form of that collector [13], we decided to use not only the exponential distribution s [14,10].…”

“…(Heap equilibrium has been the underlying assumption in some comparative analyses of garbage collection costs [3,13]. A relative heap size parameter is used as the basis for comparison of two collection algorithms: heap size is a fixed multiple of a steady-state live data amount.)…”