TY - GEN
T1 - The dynamics of backfilling
T2 - IEEE International Symposium on Workload Characterization, IISWC-2006
AU - Tsafrir, Dan
AU - Feitelson, Dror G.
PY - 2006
Y1 - 2006
N2 - Parallel job scheduling with backfilling requires users to provide runtime estimates, used by the scheduler to better pack the jobs. Studies of the impact of such estimates on performance have modeled them using a "badness factor" f ≥ 0 in an attempt to capture their inaccuracy (given a runtime r, the estimate is uniformly distributed in [r, (f + 1) · r]). Surprisingly, inaccurate estimates (f > 0) yielded better performance than accurate ones (f = 0). We explain this by a "heel and toe" dynamics that, with f > 0, cause backfilling to approximate shortest-job first scheduling. We further find the effect of systematically increasing f is V-shaped: average wait time and slowdown initially drop, only to rise later on. This happens because higher fs create bigger "holes" in the schedule (longer jobs can backfill) and increase the randomness (more long jobs appear as short), thus overshadowing the initial heel-and-toe preference for shorter jobs. The bottom line is that artificial inaccuracy generated by multiplying (real or perfect) estimates by a factor is (1) just a scheduling technique that trades off fairness for performance, and is (2) ill-suited for studying the effect of real inaccuracy. Real estimates are modal (90% of the jobs use the same 20 estimates) and bounded by a maximum (usually the most popular estimate). Therefore, when performing an evaluation, "increased inaccuracy" should translate to increased modality. Unlike multiplying, this indeed worsens performance as one would intuitively expect.
AB - Parallel job scheduling with backfilling requires users to provide runtime estimates, used by the scheduler to better pack the jobs. Studies of the impact of such estimates on performance have modeled them using a "badness factor" f ≥ 0 in an attempt to capture their inaccuracy (given a runtime r, the estimate is uniformly distributed in [r, (f + 1) · r]). Surprisingly, inaccurate estimates (f > 0) yielded better performance than accurate ones (f = 0). We explain this by a "heel and toe" dynamics that, with f > 0, cause backfilling to approximate shortest-job first scheduling. We further find the effect of systematically increasing f is V-shaped: average wait time and slowdown initially drop, only to rise later on. This happens because higher fs create bigger "holes" in the schedule (longer jobs can backfill) and increase the randomness (more long jobs appear as short), thus overshadowing the initial heel-and-toe preference for shorter jobs. The bottom line is that artificial inaccuracy generated by multiplying (real or perfect) estimates by a factor is (1) just a scheduling technique that trades off fairness for performance, and is (2) ill-suited for studying the effect of real inaccuracy. Real estimates are modal (90% of the jobs use the same 20 estimates) and bounded by a maximum (usually the most popular estimate). Therefore, when performing an evaluation, "increased inaccuracy" should translate to increased modality. Unlike multiplying, this indeed worsens performance as one would intuitively expect.
UR - http://www.scopus.com/inward/record.url?scp=48449088840&partnerID=8YFLogxK
U2 - 10.1109/IISWC.2006.302737
DO - 10.1109/IISWC.2006.302737
M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???
AN - SCOPUS:48449088840
SN - 1424405084
SN - 9781424405084
T3 - Proceedings of the 2006 IEEE International Symposium on Workload Characterization, IISWC - 2006
SP - 131
EP - 141
BT - Proceedings of the 2006 IEEE International Symposium on Workload Characterization, IISWC - 2006
Y2 - 25 October 2006 through 27 October 2006
ER -