arXiv subjects

Bernhard Scholz

Publications and source records attributed to Bernhard Scholz.

16 recordsLinked to original sources

arXiv2026

SonicDB S6: A Storage-Efficient Verkle Trie for High-Throughput Blockchains

The Ethereum state database uses Merkle Patricia Trie (MPT), which suffers from large witness proof sizes and high storage overhead. Verkle Tries have been proposed as a replacement, offering witness proofs below 150 bytes through vector commitments and Inner Product Argument aggregation. However, deploying a Verkle Trie in a high-throughput, short block-time blockchain such as Sonic, which produces a block every 300 milliseconds, introduces substantial engineering challenges related to storage efficiency, commitment computation costs, and the need to serve both live and historical state queries in real time. We present SonicDB S6, a production-grade Rust Verkle Trie database for the Sonic blockchain, which leverages its non-forking property to enable aggressive storage optimizations. Occupancy-aware node specializations, selected via an $\mathcal{O}(k n^2)$ dynamic program, reduce live storage by 97.8\%. Delta nodes that record only changed slots reduce archive storage by 95\%. Batched updates, multi-threaded commitment computation, and homomorphic Pedersen caching yield $3.2\times$ higher throughput than a persistent Geth Verkle baseline while sustaining production block-rate performance.

cs.DB↗

arXiv2025

Efficient Forkless Blockchain Databases

Operating nodes in an L1 blockchain remains costly despite recent advances in blockchain technology. One of the most resource-intensive components of a node is the blockchain database, also known as StateDB, that manages balances, nonce, code, and the persistent storage of accounts/smart contracts. Although the blockchain industry has transitioned from forking to forkless chains due to improved consensus protocols, forkless blockchains still rely on legacy forking databases that are suboptimal for their purposes. In this paper, we propose a forkless blockchain database, showing a 100x improvement in storage and a 10x improvement in throughput compared to the geth-based Fantom Blockchain client.

cs.DB↗

arXiv2025

A Fast Ethereum-Compatible Forkless Database

The State Database of a blockchain stores account data and enables authentication. Modern blockchains use fast consensus protocols to avoid forking, improving throughput and finality. However, Ethereum's StateDB was designed for a forking chain that maintains multiple state versions. While newer blockchains adopt Ethereum's standard for DApp compatibility, they do not require multiple state versions, making legacy Ethereum databases inefficient for fast, non-forking blockchains. Moreover, existing StateDB implementations have been built on key-value stores (e.g., LevelDB), which make them less efficient. This paper introduces a novel state database that is a native database implementation and maintains Ethereum compatibility while being specialized for non-forking blockchains. Our database delivers ten times speedups and 99% space reductions for validators, and a threefold decrease in storage requirements for archive nodes.

cs.DB↗

arXiv2025

Reusable Formal Verification of DAG-based Consensus Protocols

Blockchains use consensus protocols to reach agreement, e.g., on the ordering of transactions. DAG-based consensus protocols are increasingly adopted by blockchain companies to reduce energy consumption and enhance security. These protocols collaboratively construct a partial order of blocks (DAG construction) and produce a linear sequence of blocks (DAG ordering). Given the strategic significance of blockchains, formal proofs of the correctness of key components such as consensus protocols are essential. This paper presents safety-verified specifications for five DAG-based consensus protocols. Four of these protocols -- DAG-Rider, Cordial Miners, Hashgraph, and Eventual Synchronous BullShark -- are well-established in the literature. The fifth protocol is a minor variation of Aleph, another well-established protocol. Our framework enables proof reuse, reducing proof efforts by almost half. It achieves this by providing various independent, formally verified, specifications of DAG construction and ordering variations, which can be combined to express all five protocols. We employ TLA+ for specifying the protocols and writing their proofs, and the TLAPS proof system to automatically check the proofs. Each TLA+ specification is relatively compact, and TLAPS efficiently verifies hundreds to thousands of obligations within minutes. The significance of our work is two-fold: first, it supports the adoption of DAG-based systems by providing robust safety assurances; second, it illustrates that DAG-based consensus protocols are amenable to practical, reusable, and compositional formal methods.

cs.LO↗

arXiv2025

Formal Model Guided Conformance Testing for Blockchains

Modern blockchains increasingly consist of multiple clients that implement a single blockchain protocol. If there is a semantic mismatch between the protocol implementations, the blockchain can permanently split and introduce new attack vectors. Current ad-hoc test suites for client implementations are not sufficient to ensure a high degree of protocol conformance. As an alternative, we present a framework that performs protocol conformance testing using a formal model of the protocol and an implementation running inside a deterministic blockchain simulator. Our framework consists of two complementary workflows that use the components as trace generators and checkers. Our insight is that both workflows are needed to detect all types of violations. We have applied and demonstrated the utility of our framework on an industrial strength consensus protocol.

cs.CR↗

arXiv2025

Provenance Guided Rollback Suggestions

Advances in incremental Datalog evaluation strategies have made Datalog popular among use cases with constantly evolving inputs such as static analysis in continuous integration and deployment pipelines. As a result, new logic programming debugging techniques are needed to support these emerging use cases. This paper introduces an incremental debugging technique for Datalog, which determines the failing changes for a \emph{rollback} in an incremental setup. Our debugging technique leverages a novel incremental provenance method. We have implemented our technique using an incremental version of the Soufflé Datalog engine and evaluated its effectiveness on the DaCapo Java program benchmarks analyzed by the Doop static analysis library. Compared to state-of-the-art techniques, we can localize faults and suggest rollbacks with an overall speedup of over 26.9$\times$ while providing higher quality results.

cs.LO↗

arXiv2023

Julia Cloud Matrix Machine: Dynamic Matrix Language Acceleration on Multicore Clusters in the Cloud

In emerging scientific computing environments, matrix computations of increasing size and complexity are increasingly becoming prevalent. However, contemporary matrix language implementations are insufficient in their support for efficient utilization of cloud computing resources, particularly on the user side. We thus developed an extension of the Julia high-performance computation language such that matrix computations are automatically parallelized in the cloud, where users are separated from directly interacting with complex explicitly-parallel computations. We implement lazy evaluation semantics combined with directed graphs to optimize matrix operations on the fly while dynamic simulation finds the optimal tile size and schedule for a given cluster of cloud nodes. A time model prediction of the cluster's performance capacity is constructed to enable simulations. Automatic configuration of communication and worker processes on the cloud networks allow for the framework to automatically scale up for clusters of heterogeneous nodes. Our framework's experimental evaluation comprises eleven benchmarks on an fourteen node (564 CPUs) cluster in the AWS public cloud, revealing speedups of up to a factor of 5.1, with an average 74.39% of the upper bound for speedups.

cs.DC↗

arXiv2023

Cloudprofiler: TSC-based inter-node profiling and high-throughput data ingestion for cloud streaming workloads

To conduct real-time analytics computations, big data stream processing engines are required to process unbounded data streams at millions of events per second. However, current streaming engines exhibit low throughput and high tuple processing latency. Performance engineering is complicated by the fact that streaming engines constitute complex distributed systems consisting of multiple nodes in the cloud. A profiling technique is required that is capable of measuring time durations at high accuracy across nodes. Standard clock synchronization techniques such as the network time protocol (NTP) are limited to millisecond accuracy, and hence cannot be used. We propose a profiling technique that relates the time-stamp counters (TSCs) of nodes to measure the duration of events in a streaming framework. The precision of the TSC relation determines the accuracy of the measured duration. The TSC relation is conducted in quiescent periods of the network to achieve accuracy in the tens of microseconds. We propose a throughput-controlled data generator to reliably determine the sustainable throughput of a streaming engine. To facilitate high-throughput data ingestion, we propose a concurrent object factory that moves the deserialization overhead of incoming data tuples off the critical path of the streaming framework. The evaluation of the proposed techniques within the Apache Storm streaming framework on the Google Compute Engine public cloud shows that data ingestion increases from $700$ $\text{k}$ to $4.68$ $\text{M}$ tuples per second, and that time durations can be profiled at a measurement accuracy of $92$ $μ\text{s}$, which is three orders of magnitude higher than the accuracy of NTP, and one order of magnitude higher than prior work.

cs.DC↗

arXiv2021

The Choice Construct in the Souffle Language

Datalog has become a popular implementation language for solving large-scale, real-world problems, including bug finders, network analysis tools, and disassemblers. These applications express complex behaviour with hundreds of relations and rules that often require a non-deterministic choice for tuples in relations to express worklist algorithms. This work is an experience report that describes the implementation of a choice construct in the Datalog engine Souffle. With the choice construct, we can express worklist algorithms such as spanning trees in a few lines of code. We highlight the differences between rule-based choice as described in prior work, and relation-based choice introduced by this work. We show that a choice construct enables certain worklist algorithms to be computed up to 10kx faster than having no choice construct.

cs.PL↗

arXiv2019

Worst-Case Optimal Radix Triejoin

Relatively recently, the field of join processing has been swayed by the discovery of a new class of multi-way join algorithms. The new algorithms join multiple relations simultaneously rather than perform a series of pairwise joins. The new join algorithms satisfy stronger worst-case runtime complexity guarantees than any of the existing approaches based on pairwise joins -- they are worst-case optimal in data complexity. These research efforts have resulted in a flurry of papers documenting theoretical and some practical contributions. However, there is still the quest of making the new worst-case optimal join algorithms truly practical in terms of (1) ease of implementation and (2) secondary index efficiency in terms of number of indexes created to answer a query. In this paper, we present a simple worst-case optimal multi-way join algorithm called the radix triejoin. Radix triejoin uses a binary encoding for reducing the domain of a database. Our main technical contribution is that domain reduction allows a bit-interleaving of attribute values that gives rise to a query-independent relation representation, permitting the computation of multiple queries over the same relations worst-case optimally without having to construct additional secondary indexes. We also generalise the core algorithm to conjunctive queries with inequality constraints and provide a new proof technique for the worst-case optimal join result.

cs.DB↗

arXiv2019

The Economics of Smart Contracts

Ethereum is a distributed blockchain that can execute smart contracts, which inter-communicate and perform transactions automatically. The execution of smart contracts is paid in the form of gas, which is a monetary unit used in the Ethereum blockchain. The Ethereum Virtual Machine (EVM) provides the metering capability for smart contract execution. Instruction costs vary depending on the instruction type and the approximate computational resources required to execute the instruction on the network. The cost of gas is adjusted using transaction fees to ensure adequate payment of the network. In this work, we highlight the "real" economics of smart contracts. We show that the actual costs of executing smart contracts are disproportionate to the computational costs and that this gap is continuously widening. We show that the gas cost-model of the underlying EVM instruction-set is wrongly modeled. Specifically, the computational cost for the SLOAD instruction increases with the length of the blockchain. Our proposed performance model estimates gas usage and execution time of a smart contract at a given block-height. The new gas-cost model incorporates the block-height to eliminate irregularities in the Ethereum gas calculations. Our findings are based on extensive experiments over the entire history of the EVM blockchain.

cs.DC↗

arXiv2019

Provenance for Large-scale Datalog

Logic programming languages such as Datalog have become popular as Domain Specific Languages (DSLs) for solving large-scale, real-world problems, in particular, static program analysis and network analysis. The logic specifications which model analysis problems, process millions of tuples of data and contain hundreds of highly recursive rules. As a result, they are notoriously difficult to debug. While the database community has proposed several data-provenance techniques that address the Declarative Debugging Challenge for Databases, in the cases of analysis problems, these state-of-the-art techniques do not scale. In this paper, we introduce a novel bottom-up Datalog evaluation strategy for debugging: our provenance evaluation strategy relies on a new provenance lattice that includes proof annotations, and a new fixed-point semantics for semi-naive evaluation. A debugging query mechanism allows arbitrary provenance queries, constructing partial proof trees of tuples with minimal height. We integrate our technique into Souffle, a Datalog engine that synthesizes C++ code, and achieve high performance by using specialized parallel data structures. Experiments are conducted with DOOP/DaCapo, producing proof annotations for tens of millions of output tuples. We show that our method has a runtime overhead of 1.27x on average while being more flexible than existing state-of-the-art techniques.

cs.PL↗

arXiv2018

Vandal: A Scalable Security Analysis Framework for Smart Contracts

The rise of modern blockchains has facilitated the emergence of smart contracts: autonomous programs that live and run on the blockchain. Smart contracts have seen a rapid climb to prominence, with applications predicted in law, business, commerce, and governance. Smart contracts are commonly written in a high-level language such as Ethereum's Solidity, and translated to compact low-level bytecode for deployment on the blockchain. Once deployed, the bytecode is autonomously executed, usually by a %Turing-complete virtual machine. As with all programs, smart contracts can be highly vulnerable to malicious attacks due to deficient programming methodologies, languages, and toolchains, including buggy compilers. At the same time, smart contracts are also high-value targets, often commanding large amounts of cryptocurrency. Hence, developers and auditors need security frameworks capable of analysing low-level bytecode to detect potential security vulnerabilities. In this paper, we present Vandal: a security analysis framework for Ethereum smart contracts. Vandal consists of an analysis pipeline that converts low-level Ethereum Virtual Machine (EVM) bytecode to semantic logic relations. Users of the framework can express security analyses in a declarative fashion: a security analysis is expressed in a logic specification written in the \souffle language. We conduct a large-scale empirical study for a set of common smart contract security vulnerabilities, and show the effectiveness and efficiency of Vandal. Vandal is both fast and robust, successfully analysing over 95\% of all 141k unique contracts with an average runtime of 4.15 seconds; outperforming the current state of the art tools---Oyente, EthIR, Mythril, and Rattle---under equivalent conditions.

cs.PL↗

arXiv2017

SPARK: Static Program Analysis Reasoning and Retrieving Knowledge

Program analysis is a technique to reason about programs without executing them, and it has various applications in compilers, integrated development environments, and security. In this work, we present a machine learning pipeline that induces a security analyzer for programs by example. The security analyzer determines whether a program is either secure or insecure based on symbolic rules that were deduced by our machine learning pipeline. The machine pipeline is two-staged consisting of a Recurrent Neural Networks (RNN) and an Extractor that converts an RNN to symbolic rules. To evaluate the quality of the learned symbolic rules, we propose a sampling-based similarity measurement between two infinite regular languages. We conduct a case study using real-world data. In this work, we discuss the limitations of existing techniques and possible improvements in the future. The results show that with sufficient training data and a fair distribution of program paths it is feasible to deducing symbolic security rules for the OpenJDK library with millions lines of code.

cs.PL↗

arXiv2017

Optimal On The Fly Index Selection in Polynomial Time

The index selection problem (ISP) is an important problem for accelerating the execution of relational queries, and it has received a lot of attention as a combinatorial knapsack problem in the past. Various solutions to this very hard problem have been provided. In contrast to existing literature, we change the underlying assumptions of the problem definition: we adapt the problem for systems that store relations in memory, and use complex specification languages, e.g., Datalog. In our framework, we decompose complex queries into primitive searches that select tuples in a relation for which an equality predicate holds. A primitive search can be accelerated by an index exhibiting a worst-case run-time complexity of log-linear time in the size of the output result of the primitive search. However, the overheads associated with maintaining indexes are very costly in terms of memory and computing time. In this work, we present an optimal polynomial-time algorithm that finds the minimal set of indexes of a relation for a given set of primitive searches. An index may cover more than one primitive search due to the algebraic properties of the search predicate, which is a conjunction of equalities over the attributes of a relation. The index search space exhibits a exponential complexity in the number of attributes in a relation, and, hence brute-force algorithms searching for solutions in the index domain are infeasible. As a scaffolding for designing a polynomial-time algorithm, we build a partial order on search operations and use a constructive version of Dilworth's theorem. We show a strong relationship between chains of primitive searches (forming a partial order) and indexes. We demonstrate the effectiveness and efficiency of our algorithm for an in-memory Datalog compiler that is able to process relations with billions of entries in memory.

cs.DB↗

arXiv2009

Accelerating the Execution of Matrix Languages on the Cell Broadband Engine Architecture

Matrix languages, including MATLAB and Octave, are established standards for applications in science and engineering. They provide interactive programming environments that are easy to use due to their scripting languages with matrix data types. Current implementations of matrix languages do not fully utilise high-performance, special-purpose chip architectures such as the IBM PowerXCell processor (Cell), which is currently used in the fastest computer in the world. We present a new framework that extends Octave to harness the computational power of the Cell. With this framework the programmer is relieved of the burden of introducing explicit notions of parallelism. Instead the programmer uses a new matrix data-type to execute matrix operations in parallel on the synergistic processing elements (SPEs) of the Cell. We employ lazy evaluation semantics for our new matrix data-type to obtain execution traces of matrix operations. Traces are converted to data dependence graphs; operations in the data dependence graph are lowered (split into sub-matrices), scheduled and executed on the SPEs. Thereby we exploit (1) data parallelism, (2) instruction level parallelism, (3) pipeline parallelism and (4) task parallelism of matrix language programs. We conducted extensive experiments to show the validity of our approach. Our Cell-based implementation achieves speedups of up to a factor of 12 over code run on recent Intel Core2 Quad processors.

cs.PL↗