arXiv subjects

Venkat Arun

Publications and source records attributed to Venkat Arun.

13 recordsLinked to original sources

arXiv2026

DIMWIT: Directly Invoking Models Without Intensive Tuning

Resource managers typically encode what to do in a task-specific policy---a heuristic, an optimization procedure, or a model trained for the task---that must be developed before use and may need revision as the workload or system changes. We propose bypassing policy development and instead asking pretrained models to make resource-management decisions directly, given the current system state, desired outcome, and available choices. We evaluate this approach on microservice autoscaling, DRAM sizing, and video encoding, using a frontier LLM and two small decision models. On two microservice applications, a small decision model meets latency targets while using within 10% of the replicas allocated by policy-based autoscalers and within 13% of the frontier LLM. In DRAM sizing, a small decision model reduces DRAM usage by up to 27.5% relative to static provisioning while maintaining at least 95% of baseline performance. In video encoding, it completes encodes up to 4.3$\times$ faster than the frontier LLM, with an imperceptible 1.1-point decrease in Video Multi-Method Assessment Fusion (VMAF), a metric of perceptual video quality. Making individual resource-management decisions may be considerably simpler than developing a general strategy for making them.

cs.PF↗

arXiv2026

Speculative Ad-hoc Querying

Large language models (LLMs) invite substantial effort and investment, yet agentic systems can spend much of their time waiting for tools. This paper explores how a small auxiliary LLM can reduce these waits by speculatively preparing a database while the primary LLM reasons. The challenge is that the auxiliary model must anticipate the primary model's actions before the primary model takes them. Our key insight is that database systems give the speculator a structural advantage: it need not predict the exact next query. Instead, it can prepare work shared by many plausible queries. Predicting query shapes enables advance compilation, predicting data access warms caches, and predicting broader data requirements enables precomputation of superset results that subsequent queries can refine. We present Wonder, a system that exploits this advantage for analytical SQL workloads. Wonder constructs reusable superset queries, materializes their components, and rewrites subsequent queries against the resulting temporary tables. Its scheduler prioritizes foreground queries while using reasoning intervals for speculative preparation. The same engine supports humans writing SQL, using incomplete queries to guide speculation and displaying intermediate results as they type. In a replay-based evaluation of agent workflows, Wonder reduces SQL execution time by up to 31.8% and combined primary-agent reasoning and active SQL time by up to 22.6%, with approximately 6% higher recorded main-agent and prediction API cost and 10% more read I/O on the full dataset. A separate 24-participant study finds significantly lower query latency on both analytical tasks and improved time-to-answer on one, with 87.5% of participants reporting increased productivity.

cs.DB↗

arXiv2026

Robustness Starts with the Objective: Designing and Certifying Robust LLM-Generated Heuristics

LLMs are increasingly used to design heuristics for computer systems. However, most synthesis frameworks evaluate candidates by their average performance, which can mask severe regressions on specific workloads. Even adding adversarial training inputs doesn't eliminate these regressions, and percentile objectives (e.g., optimizing for the 95th percentile) overlook how severe regressions can be beyond the selected threshold. We present TAILOR, a framework that guides LLM synthesis using the Conditional Value at Risk (CVaR) of the cost difference between each candidate and an existing baseline on the same workload. This paired CVaR measures the average difference over the worst fraction of workloads. We show that a negative paired CVaR guarantees improvements in both average performance and tail optimality gap relative to the unknown optimum, without ever computing that optimum. We find that TAILOR achieves the best mean--tail trade-off among the synthesis objectives we evaluate across three case studies: Azure GPU batch admission, Azure VM packing, and server placement. Tested across seven diverse optimization benchmarks, TAILOR produces heuristics that are certified with 99% confidence to improve both average performance and tail optimality gaps over the baseline in six of them. On these six benchmarks, TAILOR also outperforms known robust, human-designed alternatives in the tail while achieving better or comparable average-performance improvement over the baseline.

cs.AI↗

arXiv2026

Agar: What-If Analysis for Microservice Applications

Microservice operators must reason about how applications would behave under changes that have not yet occurred: different workloads, resource allocations, or SLO targets. Evaluating every alternative experimentally is expensive, but predicting such counterfactuals is difficult because microservices contain many interacting components. End-to-end models must learn behavior over a combinatorial space of conditions, but composing per-component predictors causes prediction errors to accumulate. We address this challenge by changing what each component model must predict. Rather than predicting exact values, we learn simple, interpretable bounds that contain every observed behavior while remaining tight enough to support useful end-to-end conclusions. We build AGAR, a counterfactual performance analysis tool that composes these models to answer end-to-end SLO-feasibility questions. This allows us to formulate autoscaling as a counterfactual query: what is the smallest resource allocation under which an SLO is met? Our evaluation demonstrates that across three microservice applications, AGAR's bounds are actionable: its derived autoscaler reduces CPU allocation by up to 36.8% over state-of-the-art horizontal autoscalers while satisfying application SLOs. Further, AGAR is able to predict operational cost given workload traces, provide guidance on which service's startup time to optimize and localize performance bugs within each application.

cs.DC↗

arXiv2026

Assistants, Not Architects: The Role of LLMs in Networked Systems Design

Designing the architecture of modern networked systems requires navigating a large, combinatorial space of hardware, systems, and configuration choices with complex cross-layer interactions. Architects must balance competing objectives such as performance, cost, and deployability while satisfying compatibility and resource constraints, often relying on scattered rules-of-thumb drawn from benchmarks, papers, documentation, and expert experience. This raises a natural question: can large language models (LLMs) reliably perform this kind of architectural reasoning? We find that they cannot. While LLMs produce plausible configurations, they frequently miss critical constraints, encode incorrect assumptions, and exhibit ``stickiness'' to familiar patterns. A natural workaround--iterative validation via simulation or experimentation--is often prohibitively expensive at scale and, in many cases, infeasible, particularly when comparing hardware-dependent alternatives. Motivated by this gap, we present Kepler, a lightweight reasoning framework for architecture design that combines structured, expert-driven specifications with SMT-based optimization. Kepler encodes architecturally significant properties--requirements, incompatibilities, and qualitative trade-offs--about systems, hardware, and workloads as constraints, and synthesizes feasible designs that optimize user-defined objectives. It operates at an abstract level, capturing ``rules-of-thumb'' rather than detailed system behavior, enabling tractable reasoning while preserving key interactions, and provides explanations for its decisions. Through experiments and case studies, we show that Kepler uncovers interactions missed by LLMs and supports systematic, explainable design exploration.

cs.NI↗

arXiv2026

Bring Your Own Objective: Inter-operability of Network Objectives in Datacenters

Datacenter networks are currently locked in a "tyranny of the single objective". While modern workloads demand diverse performance goals, ranging from coflow completion times, per-flow fairness, short-flow latencies, existing fabrics are typically hardcoded for a single metric. This rigid coupling ensures peak performance when application and network objectives align, but results in abysmal performance when they diverge. We propose DMart, a decentralized scheduling framework that treats network bandwidth as a competitive marketplace. In DMart, applications independently encode the urgency and importance of their network traffic into autonomous bids, allowing diverse objectives to co-exist natively on the same fabric. To meet the extreme scale and sub-microsecond requirements of modern datacenters, DMart implements distributed, per-link, per-RTT auctions, without relying on ILPs, centralized schedulers, or complex priority queues. We evaluate DMart using packet-level simulations and compare it against network schedulers designed for individual metrics, e.g., pFabric and Sincronia. DMart matches the performance of specialized schedulers on their own "home turf" while simultaneously optimizing secondary metrics. Compared to pFabric and Sincronia, DMart reduces deadline misses by 2x and coflow completion times by 1.6x respectively, while matching pFabric short-flow completion times.

cs.NI↗

arXiv2025

Robust Heuristic Algorithm Design with LLMs

We posit that we can generate more robust and performant heuristics if we augment approaches using LLMs for heuristic design with tools that explain why heuristics underperform and suggestions about how to fix them. We find even simple ideas that (1) expose the LLM to instances where the heuristic underperforms; (2) explain why they occur; and (3) specialize design to regions in the input space, can produce more robust algorithms compared to existing techniques~ -- ~the heuristics we produce have a $\sim28\times$ better worst-case performance compared to FunSearch, improve average performance, and maintain the runtime.

cs.AI↗

arXiv2025

Contracts: A unified lens on congestion control robustness, fairness, congestion, and generality

Congestion control algorithms (CCAs) operate in partially observable environments, lacking direct visibility into link capacities, or competing flows. To ensure fair sharing of network resources, CCAs communicate their fair share through observable signals. For instance, Reno's fair share is encoded as $\propto 1/\sqrt{\texttt{loss rate}}$. We call such communication mechanisms \emph{contracts}. We show that the design choice of contracts fixes key steady-state performance metrics, including robustness to errors in congestion signals, fairness, amount of congestion (e.g., delay, loss), and generality (e.g., range of supported link rates). This results in fundamental tradeoffs between these metrics. Using properties of contracts we also identify design pitfalls that lead to starvation (extreme unfairness). We argue that CCA design and analysis should start with contracts to conscientiously pick tradeoffs and avoid pitfalls. We empirically validate our findings and discuss their implications on CCA design and network measurement.

cs.NI↗

arXiv2024

A Performance Verification Methodology for Resource Allocation Heuristics

Performance verification is a nascent but promising tool for understanding the performance and limitations of heuristics under realistic assumptions. Bespoke performance verification tools have already demonstrated their value in settings like congestion control and packet scheduling. In this paper, we aim to emphasize the broad applicability and utility of performance verification. To that end, we highlight the design principles of performance verification. Then, we leverage that understanding to develop a set of easy-to-follow guidelines that are applicable to a wide range of resource allocation heuristics. In particular, we introduce Virelay, a framework that enables heuristic designers to express the behavior of their algorithms and their assumptions about the system in an environment that resembles a discrete-event simulator. We demonstrate the utility and ease-of-use of Virelay by applying it to six diverse case studies. We produce bounds on the performance of classical algorithms, work stealing and SRPT scheduling, under practical assumptions. We demonstrate Virelay's expressiveness by capturing existing models for congestion control and packet scheduling, and we verify the observation that TCP unfairness can cause some ML training workloads to spontaneously converge to a state of high network utilization. Finally, we use Virelay to identify two bugs in the Linux CFS load balancer.

cs.LO↗

arXiv2021

Privid: Practical, Privacy-Preserving Video Analytics Queries

Analytics on video recorded by cameras in public areas have the potential to fuel many exciting applications, but also pose the risk of intruding on individuals' privacy. Unfortunately, existing solutions fail to practically resolve this tension between utility and privacy, relying on perfect detection of all private information in each video frame--an elusive requirement. This paper presents: (1) a new notion of differential privacy (DP) for video analytics, $(ρ,K,ε)$-event-duration privacy, which protects all private information visible for less than a particular duration, rather than relying on perfect detections of that information, and (2) a practical system called Privid that enforces duration-based privacy even with the (untrusted) analyst-provided deep neural networks that are commonplace for video analytics today. Across a variety of videos and queries, we show that Privid achieves accuracies within 79-99% of a non-private system.

cs.CR↗

arXiv2021

Throughput-Fairness Tradeoffs in Mobility Platforms

This paper studies the problem of allocating tasks from different customers to vehicles in mobility platforms, which are used for applications like food and package delivery, ridesharing, and mobile sensing. A mobility platform should allocate tasks to vehicles and schedule them in order to optimize both throughput and fairness across customers. However, existing approaches to scheduling tasks in mobility platforms ignore fairness. We introduce Mobius, a system that uses guided optimization to achieve both high throughput and fairness across customers. Mobius supports spatiotemporally diverse and dynamic customer demands. It provides a principled method to navigate inherent tradeoffs between fairness and throughput caused by shared mobility. Our evaluation demonstrates these properties, along with the versatility and scalability of Mobius, using traces gathered from ridesharing and aerial sensing applications. Our ridesharing case study shows that Mobius can schedule more than 16,000 tasks across 40 customers and 200 vehicles in an online manner.

cs.CY↗

arXiv2019

Finding Safety in Numbers with Secure Allegation Escrows

For fear of retribution, the victim of a crime may be willing to report it only if other victims of the same perpetrator also step forward. Common examples include 1) identifying oneself as the victim of sexual harassment, especially by a person in a position of authority or 2) accusing an influential politician, an authoritarian government, or ones own employer of corruption. To handle such situations, legal literature has proposed the concept of an allegation escrow: a neutral third-party that collects allegations anonymously, matches them against each other, and de-anonymizes allegers only after de-anonymity thresholds (in terms of number of co-allegers), pre-specified by the allegers, are reached. An allegation escrow can be realized as a single trusted third party; however, this party must be trusted to keep the identity of the alleger and content of the allegation private. To address this problem, this paper introduces Secure Allegation Escrows (SAE, pronounced "say"). A SAE is a group of parties with independent interests and motives, acting jointly as an escrow for collecting allegations from individuals, matching the allegations, and de-anonymizing the allegations when designated thresholds are reached. By design, SAEs provide a very strong property: No less than a majority of parties constituting a SAE can de-anonymize or disclose the content of an allegation without a sufficient number of matching allegations (even in collusion with any number of other allegers). Once a sufficient number of matching allegations exist, the join escrow discloses the allegation with the allegers' identities. We describe how SAEs can be constructed using a novel authentication protocol and a novel allegation matching and bucketing algorithm, provide formal proofs of the security of our constructions, and evaluate a prototype implementation, demonstrating feasibility in practice.

cs.CR↗

arXiv2019

RFocus: Practical Beamforming for Small Devices

To reduce transmit power, increase throughput, and improve communication range, radio systems---such as IoT sensor networks, Wi-Fi and cellular networks---benefit from the ability to direct their signals, to ensure that more of the transmitted power reaches the receiver. Many modern systems beamform with antenna arrays for this purpose. However, a radio's ability to direct its signal is fundamentally limited by its size. Unfortunately practical challenges limit the size of modern radios, and consequently, their ability to beamform. In many settings, radios on devices must be small and inexpensive; today, these settings are unable to benefit from high-precision beamforming. To address this problem, we introduce RFocus, which moves beamforming functions from the radio endpoints to the environment. RFocus includes a two-dimensional surface with a rectangular array of simple elements, each of which functions as an RF switch. Each element either lets the signal through or reflects it. The surface does not emit any power of its own. The state of the elements is set by a software controller to maximize the signal strength at a receiver, with a novel optimization algorithm that uses signal strength measurements from the receiver. The RFocus surface can be manufactured as an inexpensive thin wallpaper, requiring no wiring. This solution requires only a method to communicate received signal strengths periodically to the RFocus controller. Our prototype implementation improves the median signal strength by 10.5x, and the median channel capacity by 2.1x.

cs.NI↗