arXiv subjects

Aaron Archer

Publications and source records attributed to Aaron Archer.

3 recordsLinked to original sources

arXiv2023

Load is not what you should balance: Introducing Prequal

We present Prequal (Probing to Reduce Queuing and Latency), a load balancer for distributed multi-tenant systems. Prequal aims to minimize real-time request latency in the presence of heterogeneous server capacities and non-uniform, time-varying antagonist load. It actively probes server load to leverage the power-of-d-choices paradigm, extending it with asynchronous and reusable probes. Cutting against received wisdom, Prequal does not balance CPU load, but instead selects servers according to estimated latency and active requests-in-flight (RIF). We explore its major design features on a testbed system and evaluate it on YouTube, where it has been deployed for more than two years. Prequal has dramatically decreased tail latency, error rates, and resource use, enabling YouTube and other production systems at Google to run at much higher utilization.

cs.DC↗

arXiv2023

Practical Performance Guarantees for Pipelined DNN Inference

We optimize pipeline parallelism for deep neural network (DNN) inference by partitioning model graphs into $k$ stages and minimizing the running time of the bottleneck stage, including communication. We give practical and effective algorithms for this NP-hard problem, but our emphasis is on tackling the practitioner's dilemma of deciding when a solution is good enough. To this end, we design novel mixed-integer programming (MIP) relaxations for proving lower bounds. Applying these methods to a diverse testbed of 369 production models, for $k \in \{2, 4, 8, 16, 32, 64\}$, we empirically show that these lower bounds are strong enough to be useful in practice. Our lower bounds are substantially stronger than standard combinatorial bounds. For example, evaluated via geometric means across a production testbed with $k = 16$ pipeline stages, our MIP formulations raise the lower bound from 0.4598 to 0.9452, expressed as a fraction of the best partition found. In other words, our improved lower bounds close the optimality gap by a factor of 9.855x.

cs.LG↗

arXiv2018

Wireless coverage prediction via parametric shortest paths

When deciding where to place access points in a wireless network, it is useful to model the signal propagation loss between a proposed antenna location and the areas it may cover. The indoor dominant path (IDP) model, introduced by Wölfle et al., is shown in the literature to have good validation and generalization error, is faster to compute than competing methods, and is used in commercial software such as WinProp, iBwave Design, and CellTrace. Previously, the algorithms known for computing it involved a worst-case exponential-time tree search, with pruning heuristics to speed it up. We prove that the IDP model can be reduced to a parametric shortest path computation on a graph derived from the walls in the floorplan. It therefore admits a quasipolynomial-time (i.e., $n^{O(\log n)}$) algorithm. We also give a practical approximation algorithm based on running a small constant number of shortest path computations. Its provable worst-case additive error (in dB) can be made arbitrarily small via appropriate choices of parameters, and is well below 1dB for reasonable choices. We evaluate our approximation algorithm empirically against the exact IDP model, and show that it consistently beats its theoretical worst-case bounds, solving the model exactly (i.e., no error) in the vast majority of cases.

cs.DS↗