Mobile cloud backup applications are widely used to safeguard user data. This paper found that current cloud backup is inefficient on resource-limited mobile devices because it consumes excessive CPU resources for delta synchronization that requires intensive hash computation to identify the modified ranges of file data. To address this issue, this paper presents SolFS, an operation log versioning file system to optimize mobile cloud backup efficiency. The core idea is that if the cloud backup application knows the modified offset and length of each write since the last backup, it will be able to identify the new modified data and upload them only, avoiding data hashing throughout the entire file. SolFS proposes a series of designs to achieve this design goal. First, SolFS introduces per-file mergeable operation logging that allows each file to manage its write operation logs (i.e., offset and length) in an extent tree and merge operation logs with contiguous or overlapping modified ranges of file data. Then, SolFS proposes the operation log persistence and versioning mechanism that allows different cloud backup applications to manage their own file versions without interfering with each other. In addition, SolFS incorporates techniques such as compact log and dynamic granularity, to optimize the memory and storage overhead to the system. Finally, SolFS achieves hash-free file difference identification with the minimum extra overhead and extends the ability of cloud backup applications. In addition to the APP side, we extend the SolFS design to the cloud backup server to reduce the server-side computation overhead. Furthermore, the SolFS-enabled server integrates a new shadow file mechanism to significantly mitigate the write overhead caused by file reconstruction. The experimental results show that SolFS can significantly reduce the computational overhead of server-side by over 90% on average and the total cloud synchronization time by over 88.8% when files are updated. On the server side, the SolFS-enabled server can reduce up to 78% of file reconstruction latency.

We propose PipeANN , an on-disk graph-based approximate nearest neighbor search (ANNS) system, which significantly bridges the latency gap with in-memory ones. We achieve this by aligning the best-first search algorithm with SSD characteristics, avoiding strict compute-I/O order across search steps. Experiments show that PipeANN has 1.14 × –2.02 × search latency compared to in-memory Vamana, and 35.0% of the latency of on-disk DiskANN in billion-scale datasets, without sacrificing search accuracy. PipeANN is open-source at https://github.com/thustorage/PipeANN.

Reliable cloud-based storage systems require accurate failure prediction for Solid-State Drives (SSDs) and Hard Disk Drives (HDDs) to reduce data loss, enable proactive maintenance, support service-level reliability, and lower operational costs. In this survey, we review over 150 prior studies on storage failure prediction and related tasks, and provide a structured overview and evaluation of currently available techniques for storage failure prediction, spanning traditional statistical methods, machine learning, and deep learning approaches. We focus on device-level predictions and compare the performance, constraints, and implementation overhead of prior works in real-world scenarios. Challenges such as data imbalance, fail-slow degradation, and evolving failure patterns are discussed to identify current research gaps, such as the limited interpretability of advanced models, and the need for standardized benchmarks. Our main contribution is the introduction of structured decision frameworks that guide practitioners to choose suitable evaluation metrics, predictive models, and data preparation methods based on certain operational scenarios. These frameworks are complemented by comparative analysis of models, evaluation metrics, interpretability methods and computational overhead across deployment contexts. Our survey discusses open challenges and research directions in the domain, and offers useful insights and a structured methodology for translating research into practical deployment strategies.

Storage savings and data confidentiality are two primary yet conflicting goals in outsourced backup management. While deduplication-aware encryption has been extensively studied to make deduplication viable for encrypted data, it is incompatible with fine-grained delta and local compression for further storage savings. We present ShieldReduce, a secure outsourced storage system that aims for fine-grained shielded data reduction by applying deduplication, delta compression, and local compression to data in a trusted execution environment based on Intel SGX, so as to achieve high storage savings with security guarantees. To mitigate the I/Os of accessing base chunks for delta compression in SGX, ShieldReduce adopts bi-directional delta compression via a novel hybrid inline and offline compression design to maintain the physical locality of base chunks. It also offers a storage-prioritized offline compression mode for further storage savings and ensures crash consistency during data reduction. Evaluation on various backup workloads shows that ShieldReduce achieves significant speedups over a shielded baseline without bi-directional delta compression, while maintaining comparable storage savings to fine-grained data reduction for plain data.

Certain NAND flash-based storage devices are not equipped with dynamic random access memory (DRAM) for holding the whole mapping table, due to the constraints of chip size and the cost, such as secure digital cards. Such DRAM-less flash memory can load only a subset of frequently accessed mapping table entries into a limited-capacity, on-board static random access memory (SRAM) cache, to expedite address translation. To improve the use efficiency of the SRAM cache, this paper proposes to prefetch mapping table entries into the cache according to their locality . Then, it can minimize the number of translation page reads at the flash array caused by loading the required entries, thus improving I/O performance. Specifically, we use the indicator of runs test to reflect the locality of mapping table entries on the same translation page. When processing a missed mapping entry, it determines whether adjacent mapping entries accompanying the missed one should be loaded into the cache or not, on the basis of the runs of the target translation page. Consequently, subsequent requests requiring access to these mapping entries can be quickly responded to with the cached ones, instead of reading the target translation pages. In addition, we support cache management based on the runs test of the mapping entries to further improve the cache hit ratio. Experimental results show that our proposal can increase the hit ratio of mapping table entries by 39.4 % and reduce overall I/O latency by 23.4 % on average, in contrast to state-of-the-art schemes.

With the rise of blockchain technology, the trend toward decentralization has spread to the field of remote storage, leading to the emergence of decentralized storage as a promising model. This change is highlighted by its features of open and fair access, reduced dependence on intermediaries, and strong privacy protections. However, similar to centralized storage, the decentralization of data management presents challenges, including the separation of ownership and control, along with the need for integrity auditing on externally managed data. The current popular centralized auditing model for the mainstream cloud storage cannot be directly used for decentralized storage environments. Additionally, Homomorphic Verification Tag (HVT)-based auditing models encounter significant problems such as high computational costs and inefficient auditing processes. In response to these needs, we introduce a novel Byzantine-tolerant Lazy Auditing framework (BLA) to ensure data integrity in decentralized storage settings. A key innovation is the hierarchical architecture used: the upper level employs a simplified Practical Byzantine Fault Tolerance (PBFT) protocol to help nodes reach a consensus on data integrity audits. At the lower level, nodes are grouped into clusters based on criteria such as accessibility, organized using a block design strategy. This approach reduces unnecessary information exchange during the auditing process. It maximizes parallel processing and strengthens fault tolerance and system resilience. By distributing data, it also reduces the impact of node failures. Our theoretical analyses and empirical evaluations clearly show that BLA reduces communication complexity compared with conventional PBFT protocols. Additionally, when compared with traditional HVT-based schemes, BLA demonstrates better storage efficiency and improved computational performance, making it a viable and effective solution for data integrity auditing in decentralized storage systems.

Computing near the source of data has been proven effective in terms of energy conservation, latency improvement, and privacy preservation. With this, edge intelligence refers to local CNN inference in embedded edge devices. Because edge devices are highly resource-constrained, in practice they store CNN model(s) in external flash memory and load them (or part of) during runtime inference. However, flash memory is subject to time-related retention errors, and thus a dilemma is that without error correction, a CNN model in flash memory quickly deteriorates and becomes useless. On the other hand, strong error correction involves extra read sensing and read retrying. Due to the concern of reliability, the lineup of embedded serial flash memory offers low-density solutions only. In this study, we investigate graceful degradation of CNN models in uncorrected high-density flash storage for edge intelligence. Inspired by the observation that in popular CNN models, the majority of CNN weight parameters share a few monotonic bit patterns in their high-order bits, we propose mapping the frequent bit patterns to low flash-cell voltage levels for protection against retention errors. Furthermore, as CNN layers show different levels of tolerance to bit errors, we also propose using adaptive cell-bit density to provide non-uniform error protection among layers. We conducted experiments on popular CNN designs, including VGG16, ResNet50, and InceptionV3, under realistic effects of flash aging. Results show that with prior methods, the inference accuracy of ResNet50 degrades to 70.7% in just three months; by contrast, with our approach, the degradation is less than 1% and the accuracy remains at 91.6% after one year of retention.

To explore the opportunity of expanding the page cache using disaggregated memory for file-backed pages, this study presents BalloonStasher, an RDMA-based disaggregated memory for data-intensive applications. Utilizing the ephemeral nature of the page cache, BalloonStasher dynamically adapts to the changing page cache demands of multiple clients. BalloonStasher supports one-sided RDMA-based memory pooling or two-sided RDMA-based memory sharing when using memory nodes. Additionally, it also supports peer memory mode, which utilizes the idle memory of peer nodes. Our extensive performance study compares the benefits and limitations of the three cache modes, and shows that BalloonStasher can mitigate the memory underutilization problem and improve performance of data-intensive applications by a large margin.

Approximate nearest neighbor search (ANNS) in high-dimensional vector spaces has become increasingly crucial in database and machine learning applications. Most previous ANNS algorithms require TB-scale memory to store indices of billion-scale datasets, making their deployment extremely expensive for high-performance ANNS services. The emerging SmartSSD technology offers an opportunity to achieve scalable ANNS via near data processing (NDP). However, there remain several challenges to directly adopt existing ANNS algorithms on multiple SmartSSDs. In this paper, we present SmartANNS, a SmartSSD-empowered billion-scale ANNS solution based on a hierarchical indexing methodology. We propose several novel designs to achieve near-linear scaling with multiple SmartSSDs. First, we propose a “host CPUs + SmartSSDs” cooperative architecture incorporated with hierarchical indices to significantly reduce data accesses and computations on SmartSSDs. Second, we propose dynamic task scheduling based on optimized data layout to achieve both load balancing and data reusing for multiple SmartSSDs. Third, we further propose a learning-based shard pruning algorithm to eliminate unnecessary computations on SmartSSDs. We implement SmartANNS using Samsung’s commercial SmartSSDs. Experimental results show that SmartANNS can improve query per second (QPS) by up to 10.7 × compared with the state-of-the-art SmartSSD-based ANNS solution–CSDANNS. Moreover, SmartANNS can achieve near-linear performance scalability for large-scale datasets using multiple SmartSSDs.

The growing popularity of encrypted container images in registries poses unique challenges for storage management due to the necessity for deduplication amidst rising image volumes. Traditional deduplication struggles with encrypted content, which inherently disguises duplicate data as distinct due to its randomized nature. Current advanced methods tackle this issue by decompressing images and applying message-locked encryption (MLE). However, these techniques face considerable challenges. Minor content changes can impair deduplication effectiveness, and decompressing layers increases storage requirements. Furthermore, this process negatively impacts both the speed at which users access the images and the overall system throughput. We propose SimEnc, a high-performance and secure deduplication system for encrypted container images by exploiting multiple similarity spaces. SimEnc pioneers the integration of semantic hashing with MLE to effectively parse semantic relationships across layers, thereby increasing deduplication efficacy. This system incorporates a rapid selection mechanism for similarity spaces, offering enhanced flexibility over previous models that relied on full decompression. By adopting Huffman decoding to navigate new similarity spaces, SimEnc not only improves deduplication ratios but also enhances overall performance. Our experimental results demonstrate that SimEnc substantially reduces storage needs by up to 261.7% compared to encrypted serverless platforms and by 54.2% against plaintext registries, while also delivering superior pull latency metrics.