Search for: All records

This paper introduces Mako, a highly available, high- throughput, and horizontally scalable transactional key-value store. Mako performs strongly consistent geo-replication to maintain availability despite entire datacenter failures, uses multi-core machines for fast serializable transaction process- ing, and shards data to scale out. To achieve these properties, especially to overcome the overheads of distributed transac- tions in geo-replicated settings, Mako decouples transaction execution and replication. This enables Mako to run transactions speculatively and very fast, and replicate transactions in the background to make them fault-tolerant. The key innovation in Mako is the use of two-phase commit (2PC) speculatively to allow distributed transactions to proceed without having to wait for their decisions to be replicated, while also preventing unbounded cascading aborts if shards fail prior to the end of replication. Our experimental evaluation on Azure shows that Mako processes 3.66M TPC-C transactions per second when data is split across 10 shards, each of which runs with 24 threads. This is an 8.6×higher throughput than state-of-the-art systems optimized for geo-replication. 

This paper introduces Mako, a highly available, highthroughput, and horizontally scalable transactional key-value store. Mako performs strongly consistent geo-replication to maintain availability despite entire datacenter failures, uses multi-core machines for fast serializable transaction processing, and shards data to scale out. To achieve these properties, especially to overcome the overheads of distributed transactions in geo-replicated settings, Mako decouples transaction execution and replication. This enables Mako to run transactions speculatively and very fast, and replicate transactions in the background to make them fault-tolerant. The key innovation in Mako is the use of two-phase commit (2PC) speculatively to allow distributed transactions to proceed without having to wait for their decisions to be replicated, while also preventing unbounded cascading aborts if shards fail prior to the end of replication. Our experimental evaluation on Azure shows that Mako processes 3.66M TPC-C transactions per second when data is split across 10 shards, each of which runs with 24 threads. This is an 8.6× higher throughput than state-of-the-art systems optimized for geo-replication. 

Abstract Near the ends of their lives, supernova remnants (SNRs) enter a “radiative phase,” when efficient cooling of the postshock gas slows expansion. Understanding SNR evolution at this stage is crucial for estimating feedback in galaxies, as SNRs are expected to release energy and momentum into the interstellar medium near the ends of their lives. A standard prediction of SNR evolutionary models is that the onset of the radiative stage precipitates the formation of a dense shell behind the forward shock. In Paper I, we showed that such shell formation yields detectable nonthermal radiation from radio toγ-rays, most notably emission brightening by nearly 2 orders of magnitude. However, there remains no observational evidence for such brightening, suggesting that this standard prediction needs to be investigated. In this paper, we perform magnetohydrodynamic simulations of SNR evolution through the radiative stage, including cosmic rays (CRs) and magnetic fields to assess their dynamical roles. We find that both sources of nonthermal pressure impede shell formation, reducing shell densities by a factor of a few to more than an order of magnitude. We also use a self-consistent model of particle acceleration to estimate the nonthermal emission from these modified SNRs and demonstrate that, for reasonable CR acceleration efficiencies and magnetic field strengths, the nonthermal signatures of shell formation can all but disappear. We therefore conclude that the absence of observational signatures of shell formation represents strong evidence that nonthermal pressures from CRs and magnetic fields play a critical dynamical role in late-stage SNR evolution. 

Colloidal semiconductor nanocrystals (NCs) have emerged as promising candidates for developing solutionprocessable optical gain media with potential applications in integrated photonic circuits and lasers. However, the deployment of NCs in these technologies has been hindered by the nonradiative Auger recombination of multiexciton states, which shortens the optical gain lifetime and reduces its spectral range. Here, we demonstrate that these limitations can be overcome by using giant colloidal quantum shells (g-QSs), comprising a quantum-confined CdSe shell grown over a large (∼14 nm) CdS bulk core. Such bulk-nanoscale architecture minimizes exciton− exciton interactions, leading to suppressed Auger recombination and one of the broadest gain bandwidths reported for colloidal nanomaterials, spanning energies both above and, remarkably, below the bandgap. Ultrafast transient absorption and photoluminescence measurements demonstrate that the high-energy portion of optical gain arises from states containing more than 15 excitons per particle, while the unusual sub-bandgap gain behavior results from an Auger-assisted radiative recombination, a mechanism that has traditionally been viewed as a loss pathway. Collectively, these results reveal a unique gain regime associated with bulk-nanocrystal hybrid systems, which offers a promising path toward solution-processable light sources. 

Dynamic max-min fair allocation (DMMF) is a simple and popular mechanism for the repeated allocation of a shared resource among competing agents: in each round, each agent can choose to request or not for the resource, which is then allocated to the requesting agent with the least number of allocations received till then. Recent work has shown that under DMMF, a simple threshold-based request policy enjoys surprisingly strong robustness properties, wherein each agent can realize a significant fraction of her optimal utility irrespective of how other agents' behave. While this goes some way in mitigating the possibility of a 'tragedy of the commons' outcome, the robust policies require that an agent defend against arbitrary (possibly adversarial) behavior by other agents. This however may be far from optimal compared to real world settings, where other agents are selfish optimizers rather than adversaries. Therefore, robust guarantees give no insight on how agents behave in an equilibrium, and whether outcomes are improved under one. Our work aims to bridge this gap by studying the existence and properties of equilibria under DMMF. To this end, we first show that despite the strong robustness guarantees of the threshold based strategies,no Nash equilibrium existswhen agents participate in DMMF, each using some fixed threshold-based policy. On the positive side, however, we show that for the symmetric case, a simple data-driven request policy guarantees that no agent benefits from deviating to a different fixed threshold policy. In our proposed policy agents aim to match the historical allocation rate with a vanishing drift towards the rate optimizing overall welfare for all users. Furthermore, the resulting equilibrium outcome can be significantly better compared to what follows from the robustness guarantees. Our results are built on a complete characterization of the steady-state distribution under DMMF, as well as new techniques for analyzing strategic agent outcomes under dynamic allocation mechanisms; we hope these may prove of independent interest in related problems.