Search for: All records

Cloud systems constantly experience changes. Unfortunately, these changes often introduce regression failures, breaking the same features or functionalities repeatedly. Such failures disrupt cloud availability and waste developers' efforts in re-investigating similar incidents. In this position paper, we argue that regression failures can be effectively prevented by enforcing low-level semantics, a new class of intermediate rules empirically inferred from past incidents, yet capable of offering partial correctness guarantees. Our experience shows that such rules are valuable to strengthen system correctness guarantees and expose new bugs. 

Satellite images using multiple wavelength channels provide crucial measurements over large areas, aiding the understanding of pollution generation and transport. However, these images often contain missing data due to cloud cover and algorithm limitations. In this paper, we introduce a novel method for interpolating missing values in satellite images by incorporating pollution transport dynamics influenced by wind patterns. Our approach utilizes a fundamental physics equation to structure the covariance of missing data, improving accuracy by considering pollution transport dynamics. To address computational challenges associated with large datasets, we implement a gradient ascent algorithm. We demonstrate the effectiveness of our method through a case study, showcasing its potential for accurate interpolation in high-resolution, spatio-temporal air pollution datasets. 

Polymers are thermally insulating due to randomly oriented molecular chains, limiting their effectiveness in thermal management. However, when processed into nanofibers, polymers can exhibit significantly higher thermal conductivity, primarily due to enhanced internal structures such as crystallinity and molecular alignment. Characterizing these structural parameters at the single nanofiber level remains a challenge, limiting understanding of thermal transport mechanisms. Here, we investigate the relationship between internal structure and thermal conductivity of single polyethylene oxide (PEO) nanofibers fabricated from near-field electrospinning (NFES). By varying molecular weight and concentration of PEO, their impact on thermal conductivity and internal structure are examined. Crystallinity is examined using conventional Raman spectroscopy, while molecular orientation is assessed through polarized Raman and polarized FTIR spectroscopy. Results reveal that enhanced thermal conductivity in PEO nanofibers is primarily attributed to increased molecular orientation. A maximum thermal conductivity of 2.7 W/m·K is achieved in PEO nanofibers, representing a notable improvement over bulk PEO (0.2 W/m·K). These findings demonstrate the potential of structurally engineered PEO nanofibers for thermal applications including electronic packaging and thermal interface materials. Further, the approach presented in this work can provide a framework for exploring thermal transport mechanisms in other polymer systems. 

Wide-area scaling trends require new approaches to Internet Protocol (IP) lookup, enabled by modern networking chips such as Intel Tofino [35], AMD Pensando [2], and Nvidia BlueField [55], which provide substantial ternary content-addressable memory (TCAM) and static random-access memory (SRAM). However, designing and evaluating scalable algorithms for these chips is challenging due to hardware-level constraints. To address this, we introduce the CRAM (CAM+RAM) lens, a framework that combines a formal model for evaluating algorithms on modern network processors with a set of optimization idioms. We demonstrate the effectiveness of CRAM by designing and evaluating three new IP lookup schemes: RESAIL, BSIC, and MASHUP. RESAIL enables Tofino-2 to scale to 2.25 million IPv4 prefixes-- likely sufficient for the next decade--while a pure TCAM approach supports only 250k prefixes, just 27% of the current global IPv4 routing table. Similarly, BSIC scales to 390k IPv6 prefixes on Tofino-2, supporting 3.2 times as many prefixes as a pure TCAM implementation. In contrast, existing state-of-the-art algorithms, SAIL [83] for IPv4 and HI-BST [65] for IPv6, scale to considerably smaller sizes on Tofino-2.