This paper presents PIP (Prediction-Informed Power), a power control framework for general-purpose compute servers. PIP introduces two key innovations: (1) a machine learning-based power model that predicts the impact of hypothetical CPU throttling actions before execution, and (2) a prediction-informed control loop that selects CPU configurations to maximize performance and power utilization based on these predictions. By leveraging finegrained runtime CPU metrics, PIP can accurately estimate counterfactual power usage, allowing the control system to align power demand with the budget more quickly. Unlike traditional reactive approaches, PIP maintains effective control under frequent budget fluctuations, achieving safe oversubscription by up to 70%. Our evaluation on diverse application workloads, none of which are included in the model's training set, shows that PIP yields up to a 3.2× speedup over a state-of-the-art feedback-based system for single-application runs, and up to a 3.4× speedup for multi-application scenarios under power constraints.

Power management has become important in data centers. Since data center workloads are often dynamic, it is common practice to conserve energy by scaling resources up or down to match the workload. And since data centers often oversubscribe the power delivery infrastructure, operators can add power capping on top of these workload-proportional systems to adjust to available power. We find that this combination leaves performance on the table and provides only a limited power control range. Instead, we argue for power-adaptive systems that attempt to make the best use of the available power budget. To illustrate this approach, we built PASS, a power-adaptive storage system. PASS considers the interactions between different system components, including software and hardware, when making its power management decisions. For example, when under the same power constraint, PASS achieves 3–25× better throughput on filebench workloads than Intel's SPDK storage stack with Google Thunderbolt, a state of the art power capping system.

In multi-container environments, applications oftentimes experience unexpected performance fluctuations due to undesirable interference among applications. Synchronization such as locks has been targeted as one of the reasons but still remains an uncontrolled resource while a large set of locks are still shared across applications. In this paper, we demonstrate that this lack of lock scheduling incurs significant real-world problems including performance unfairness and interference among applications. To address this problem, we propose a new synchronization design with an embedded scheduling capability, called CFL (Completely Fair Locking). CFL fairly distributes a fair amount of lock occupation time to applications considering their priorities and cgroup information. For scalability, CFL also considers the NUMA topology in the case of NUMA machines. Experimental results demonstrate that CFL significantly improves performance fairness while achieving comparable or sometimes even superior performance to state-of-the-art locks.

Filesystem fragmentation has been one of the primary reasons for computer systems to get slower over time. However, there have been rapid changes in modern storage systems over the past decades, and modern storage devices such as solid state drives have different mechanisms to access data, compared with traditional rotational ones. In this article, we revisit filesystem fragmentation on modern computer systems from both performance and fairness perspectives. According to our extensive experiments, filesystem fragmentation not only degrades I/O performance of modern storage devices, but also incurs various problems related to I/O fairness, such as performance interference. Unfortunately, conventional defragmentation tools are designed primarily for hard disk drives and thus generate an unnecessarily large amount of I/Os for data migration. To mitigate such problems, this article present FragPicker, a new defragmentation tool for modern storage devices. FragPicker analyzes the I/O behaviors of each target application and defragments only necessary pieces of data whose migration can contribute to performance improvement, thereby effectively minimizing the I/O amount for defragmentation. Our evaluation with YCSB workload-C shows FragPicker reduces the total amount of I/O for defragmentation by around 66% and the elapsed time by around 84%, while showing a similar level of defragmentation effect.

While gaining popularity in mobile devices, F2FS, a flash-friendly variation of log-structured file system, reveals three drawbacks: segment cleaning overhead, metadata update overhead, and file fragmentation, which becomes conspicuous under random update workloads. This paper suggests for the first time to leverage the address-remap technique in flash storage to remedy such pitfalls in F2FS. Our approach can, while preserving the benefit of log-structured writes, achieve the eventual effect of in-place update, completely preventing three drawbacks of F2FS. It can thus significantly outperform ext4 as well as vanilla F2FS under random update workloads. Armed with another write mode, F2FS will become competitive for a wider range of applications.

File fragmentation has been widely studied for several decades due to its detrimental effects on I/O activities. However, most of the previous research focuses on its performance aspect in a single application. In this paper, we analyze the effect of fragmentation on I/O control in a consolidated system where multiple applications run simultaneously. Our evaluation demonstrates that all of the weight-based I/O control mechanisms supported by the Linux kernel fail to achieve fair I/O sharing for different reasons when they meet fragmentation. Also, we show that defragmentation can promptly antidote such failures by preventing request splitting and device-level resource conflicts.

Virtualization technology has enabled server consolidation where multiple servers are co-located on a single physical machine to improve resource utilization. In such systems, proportional I/O sharing is critical to meet the SLO (Service-Level Objectives) of the applications running in each virtual instance. However, previous studies focus on block-level I/O proportionality without considering the upper-layer I/O caches, which handle I/O requests on behalf of the underlying storage devices, thereby failing to achieve application-level proportional I/O sharing. To overcome this limitation, we propose a new cache management scheme, Weight-aware Cache (WaC), which reflects the I/O weights on cache allocation and reclamation. Specifically, WaC prioritizes higher-weighted applications in the lock acquisition process of cache allocation by re-ordering the lock waiting queue based on I/O weight. Additionally, WaC keeps the number of cache entries of each application proportional to its I/O weight, through weight-aware cache reclamation. To verify the efficacy of our scheme, we implement and evaluate WaC on both the page cache and bcache. The experimental results demonstrate that our scheme improves I/O proportionality with negligible overhead in various cases.

File fragmentation has been widely studied for several decades because it negatively influences various I/O activities. To eliminate fragmentation, most defragmentation tools migrate the entire content of files into a new area. Unfortunately, such methods inevitably generate a large amount of I/Os in the process of data migration. For this reason, the conventional tools (i) cause defragmentation to be time-consuming, (ii) significantly degrade the performance of co-running applications, and (iii) even curtail the lifetime of modern storage devices. Consequently, the current usage of defragmentation is very limited although it is necessary.