Package jline.api.da

Class Da_cacheqn_retrieval

java.lang.Object
jline.api.da.Da_cacheqn_retrieval
public class Da_cacheqn_retrieval extends Object
Decomposition-aggregation driver for a CLOSED integrated cache-queueing model whose Cache node has a delayed-hit retrieval system (Cache.setRetrievalSystem). Java port of matlab/src/api/da/da_cacheqn_retrieval.m. Adapts the integrated cacheqn decomposition: the cache is relabelled as a ClassSwitch (read -> hit / miss) and the finite-population delayed-hit coalescing is made to emerge from the closed AMVA by giving the retrieval (fetch) station a load-dependent COALESCING service rate lldscaling(k)=k/d(k), d(k)=n_eff*(1-(1-1/n_eff)^k), n_eff=nitems-totalcap. The distinct-fetch throughput saturates at 1/F and the coalescing benefit emerges from the finite population. LIMITATIONS (EXPERIMENTAL - see the MATLAB reference): hitprob/missprob are the true cache probabilities (hit=P(cached), miss=1-hit) and accurate; the delayed fraction is NOT recovered (delayedprob=0, folds into miss); the coalescing throughput benefit is captured only in direction and understated; single fetch station (single-backend) only. NO CLOSED-RETRIEVAL EXAMPLE ships in the suite (every retrieval_* example is OPEN). On a hand-built closed model this driver readily hits a reducible / singular routing and a zero read-rate denominator (NaN read rate), so a plain closed model can return an empty or unreliable table. Treat the closed path as experimental and validate against LDES. The C++ SolverMVA (line-mp) deliberately REFUSES this closed path by name; only the OPEN retrieval analyzer is ported there. Verified 2026-07-24 that the open path matches across codebases while the closed path lacks a clean, exampled reference.