arXiv subjects

Zsolt István

Publications and source records attributed to Zsolt István.

5 recordsLinked to original sources

arXiv2025

Reflex: Faster Secure Collaborative Analytics via Controlled Intermediate Result Size Disclosure

Secure Multi-Party Computation (MPC) enables collaborative analytics without exposing private data. However, OLAP queries under MPC remain prohibitively slow due to oblivious execution and padding of intermediate results with filler tuples. We present Reflex, the first framework that enables configurable trimming of intermediate results across different query operators -- joins, selections, and aggregations -- within full query plans. At its core is the Resizer operator, which can be inserted between any oblivious operators to selectively remove filler tuples under MPC using user-defined probabilistic strategies. To make privacy trade-offs interpretable, we introduce a new metric that quantifies the number of observations an attacker would need to infer the true intermediate result sizes. Reflex thus makes the performance-privacy space of secure analytics navigable, allowing users to balance efficiency and protection. Experiments show substantial runtime reductions while maintaining quantifiable privacy guarantees.

cs.DB↗

arXiv2025

OffRAC: Offloading Through Remote Accelerator Calls

Modern applications increasingly demand ultra-low latency for data processing, often facilitated by host-controlled accelerators like GPUs and FPGAs. However, significant delays result from host involvement in accessing accelerators. To address this limitation, we introduce a novel paradigm we call Offloading through Remote Accelerator Calls (OffRAC), which elevates accelerators to first-class compute resources. OffRAC enables direct calls to FPGA-based accelerators without host involvement. Utilizing the stateless function abstraction of serverless computing, with applications decomposed into simpler stateless functions, offloading promotes efficient acceleration and distribution of computational loads across the network. To realize this proposal, we present a prototype design and implementation of an OffRAC platform for FPGAs that assembles diverse requests from multiple clients into complete accelerator calls with multi-tenancy performance isolation. This design minimizes the implementation complexity for accelerator users while ensuring isolation and programmability. Results show that the OffRAC approach reduces the latency of network calls to accelerators down to approximately 10.5 us, as well as sustaining high application throughput up to 85Gbps, demonstrating scalability and efficiency, making it compelling for the next generation of low-latency applications.

cs.NI↗

arXiv2024

Benchmarking Analytical Query Processing in Intel SGXv2

Trusted Execution Environments (TEEs), such as Intel's Software Guard Extensions (SGX), are increasingly being adopted to address trust and compliance issues in the public cloud. Intel SGX's second generation (SGXv2) addresses many limitations of its predecessor (SGXv1), offering the potential for secure and efficient analytical cloud DBMSs. We assess this potential and conduct the first in-depth evaluation study of analytical query processing algorithms inside SGXv2. Our study reveals that, unlike SGXv1, state-of-the-art algorithms like radix joins and SIMD-based scans are a good starting point for achieving high-performance query processing inside SGXv2. However, subtle hardware and software differences still influence code execution inside SGX enclaves and cause substantial overheads. We investigate these differences and propose new optimizations to bring the performance inside enclaves on par with native code execution outside enclaves.

cs.DB↗

arXiv2020

Towards Improving the Performance of BFT Consensus For Future Permissioned Blockchains

Permissioned Blockchains are increasingly considered in enterprise use-cases, many of which do not require geo-distribution, or even disallow it due to legislation. Examples include country-wide networks, such as Alastria, or those deployed using cloud-based platforms such as IBM Blockchain Platform. We expect these blockchains to eventually run in environments with high bandwidth and low latency modern networks, as well as, advanced programmable hardware accelerators in servers. Even though there is renewed interest in BFT consensus algorithms with various proposals targeting Permissioned Blockchains, related work does not optimize for fast networks and does not incorporate hardware accelerators -- we make the case that doing so will pay off in the long run. To this end, we re-implemented the seminal PBFT algorithm in a way that allows us to measure different configurations of the protocol. Through this we explore the benefits of various common optimization strategies and show that the protocol is unlikely to saturate more than 10Gbps networks without relying on specialized hardware-based offloading. We discuss two concrete ways in which the cost of consensus in Permissioned Blockchains could be reduced in high speed networking environments, namely, offloading to SmartNICs and implementing the protocol on standalone FPGAs.

cs.DC↗

arXiv2020

StreamChain: Rethinking Blockchain for Datacenters

Permissioned blockchains promise secure decentralized data management in business-to-business use-cases. In contrast to Bitcoin and similar public blockchains which rely on Proof-of-Work for consensus and are deployed on thousands of geo-distributed nodes, business-to-business use-cases (such as supply chain management and banking) require significantly fewer nodes, cheaper consensus, and are often deployed in datacenter-like environments with fast networking. However, permissioned blockchains often follow the architectural thinkining behind their WAN-oriented public relatives, which results in end-to-end latencies several orders of magnitude higher than necessary. In this work, we propose StreamChain, a permissioned blockchain design that eliminates blocks in favor of processing transactions in a streaming fashion. This results in a drastically lower latency without reducing throughput or forfeiting reliability and security guarantees. To demonstrate the wide applicability of our design, we prototype StreamChain based on the Hyperledger Fabric, and show that it delivers latency two orders of magnitude lower than Fabric, while sustaining similar throughput. This performance makes StreamChain a potential alternative to traditional databases and, thanks to its streaming paradigm, enables further research around reducing latency through relying on modern hardware in datacenters.

cs.DC↗