arXiv subjects

Claudia Fohry

Publications and source records attributed to Claudia Fohry.

5 recordsLinked to original sources

arXiv2026

Protecting Futures against Silent Data Corruption -- Efficient Task Replication for Dynamic Data Dependencies

As the size of computational problems grows, so does the likelihood of Silent Data Corruptions (SDCs). A common defense is replication, where the computation is repeated and correct results are determined by majority voting. Asynchronous Many-Task (AMT) runtimes are generally well suited for this approach, since the inputs and outputs of task replicas can be compared, and the tasks can be recomputed if necessary. Most existing SDC protection schemes assume static tasks and dependencies. Dynamic settings are more challenging, especially in clusters, since the tasks/data must be tracked for the comparisons. This paper considers a particularly dynamic setting with task spawning at runtime, task communication through C++11-like promises/futures, conditional touches, and cluster-wide load balancing via work-first work stealing. We propose an approach that closely couples original and replica computations by cross-validating all outgoing effects when interacting with the runtime system. The approach selectively recomputes affected tasks only. We implemented the approach in the ItoyoriFBC runtime system and conducted preliminary experiments with Fibonacci and emulated $\mathcal{H}$-matrix LU decomposition benchmarks. Results show a factor of less than two increase of failure-free running times, despite full replication, which is mainly due to improved opportunities for load balancing resulting from the higher number of tasks. The overhead for failure correction was about 0.5% of the overall running time per SDC.

cs.DC↗

arXiv2026

Silent Data Corruption Protection through Efficient Task Replication

The trend of increasing cluster sizes of supercomputers leads to a growing susceptibility to Silent Data Corruption (SDC) that can invalidate program results. A common strategy for SDC protection is replication, where the computation is repeated, and the correct result is determined as the one that is the same in at least two different computations. Applying replication to Asynchronous Many-Task (AMT) runtimes on clusters is challenging due to dynamic task spawning and work stealing, which complicate the identification of replicated tasks. To address the challenge, this paper introduces a novel replication scheme that detects and corrects SDCs for nested fork-join programs. Briefly stated, our approach replicates the computation and records the task tree. Upon a mismatch in the final result, it traverses the tree top-down to identify all corrupted tasks that could have impacted the final result. Recovery is then performed by recomputing these tasks, while the results of correct child tasks are reused. We demonstrate our implementation within a variant of the Itoyori cluster AMT runtime. Our experimental results suggest that the time to identify and reprocess the affected tasks is negligible. The paper concludes by discussing the adaptability of our scheme to tasks that cooperate through futures.

cs.DC↗

arXiv2026

User Experiences with MPI RMA and ULFM in a Resilient Key-Value Store Implementation

As hardware failures such as node losses become increasingly common, MPI programmers may want to save vulnerable data in a resilient store. While third-party storage solutions such as Redis or the Hazelcast IMap exist, a tailored, MPI-based store may be easier to integrate and can be optimized for particular application needs. This paper considers the implementation of such a store, which is intended as a component in a resilient task-based runtime system written in MPI. The store holds redundant data copies as key-value pairs in the main memories of multiple processes. Since store access operations, such as reads and writes, are naturally one-sided, we implemented the store with passive target MPI RMA functions. Process aborts are detected with the user-level failure mitigation (ULFM) extension of Open MPI. After failures, the program recovers on the surviving processes and continues with the intact data copies. Our implementation proved difficult, since several proposed ULFM functionalities for RMA have not yet been implemented. Even assuming their existence, we think that the programming task could be simplified. This paper describes our experiences, lists functionalities that we missed, and explains a workaround that we adopted in our implementation.

cs.DC↗

arXiv2024

Scoping Review of Active Learning Strategies and their Evaluation Environments for Entity Recognition Tasks

We conducted a scoping review for active learning in the domain of natural language processing (NLP), which we summarize in accordance with the PRISMA-ScR guidelines as follows: Objective: Identify active learning strategies that were proposed for entity recognition and their evaluation environments (datasets, metrics, hardware, execution time). Design: We used Scopus and ACM as our search engines. We compared the results with two literature surveys to assess the search quality. We included peer-reviewed English publications introducing or comparing active learning strategies for entity recognition. Results: We analyzed 62 relevant papers and identified 106 active learning strategies. We grouped them into three categories: exploitation-based (60x), exploration-based (14x), and hybrid strategies (32x). We found that all studies used the F1-score as an evaluation metric. Information about hardware (6x) and execution time (13x) was only occasionally included. The 62 papers used 57 different datasets to evaluate their respective strategies. Most datasets contained newspaper articles or biomedical/medical data. Our analysis revealed that 26 out of 57 datasets are publicly accessible. Conclusion: Numerous active learning strategies have been identified, along with significant open questions that still need to be addressed. Researchers and practitioners face difficulties when making data-driven decisions about which active learning strategy to adopt. Conducting comprehensive empirical comparisons using the evaluation environment proposed in this study could help establish best practices in the domain.

cs.CL↗

arXiv2021

Checkpointing and Localized Recovery for Nested Fork-Join Programs

While checkpointing is typically combined with a restart of the whole application, localized recovery permits all but the affected processes to continue. In task-based cluster programming, for instance, the application can then be finished on the intact nodes, and the lost tasks be reassigned. This extended abstract suggests to adapt a checkpointing and localized recovery technique that has originally been developed for independent tasks to nested fork-join programs. We consider a Cilk-like work stealing scheme with work-first policy in a distributed memory setting, and describe the required algorithmic changes. The original technique has checkpointing overheads below 1% and neglectable costs for recovery, we expect the new algorithm to achieve a similar performance.

cs.DC↗