Yu Gan, Mingyu Liang, Sundar Dev, David Lo, Christina Delimitrou
{"title":"Enabling Practical Cloud Performance Debugging with Unsupervised Learning","authors":"Yu Gan, Mingyu Liang, Sundar Dev, David Lo, Christina Delimitrou","doi":"10.1145/3544497.3544503","DOIUrl":"https://doi.org/10.1145/3544497.3544503","url":null,"abstract":"Abstract-Cloud applications are increasingly shifting from large monolithic services to complex graphs of loosely-coupled microservices. Despite their benefits, microservices are prone to cascading performance issues, and can lead to prolonged periods of degraded performance. We present Sage, a machine learning-driven root cause analysis system for interactive cloud microservices that is both accurate and practical. We show that Sage correctly identifies the root causes of performance issues across a diverse set of microservices and takes action to address them, leading to more predictable, performant, and efficient cloud systems.","PeriodicalId":38935,"journal":{"name":"Operating Systems Review (ACM)","volume":"56 1","pages":"34 - 41"},"PeriodicalIF":0.0,"publicationDate":"2022-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44298865","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Pharos","authors":"Srinivas Vippagunta, Ken Finnigan, K. Pusukuri","doi":"10.1145/3544497.3544505","DOIUrl":"https://doi.org/10.1145/3544497.3544505","url":null,"abstract":"Observability is a necessary capability of modern distributed systems as it allows us to gain actionable insights about reliability, availability, performance, etc., of the system.","PeriodicalId":38935,"journal":{"name":"Operating Systems Review (ACM)","volume":"56 1","pages":"51 - 54"},"PeriodicalIF":0.0,"publicationDate":"2022-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"64052445","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Muhammed Ugur, Cheng Jiang, Alex Erf, Tanvir Ahmed Khan, Baris Kasikci
{"title":"One Profile Fits All","authors":"Muhammed Ugur, Cheng Jiang, Alex Erf, Tanvir Ahmed Khan, Baris Kasikci","doi":"10.1145/3544497.3544502","DOIUrl":"https://doi.org/10.1145/3544497.3544502","url":null,"abstract":"Modern data center applications have multi-megabyte instruc- tion footprints that easily exhaust on-chip cache structures, which typically have a size of only a couple hundred kilobytes. Consequently, today's data center applications suffer from frequent Instruction cache (I-cache) and Instruction Transla- tion Lookaside Buffer (I-TLB) misses, causing performance losses worth millions of dollars. To make matters worse, the multi-megabyte instruction footprint of the Linux kernel pre- cipitates an undue burden on the performance of data center applications.","PeriodicalId":38935,"journal":{"name":"Operating Systems Review (ACM)","volume":"56 1","pages":"26 - 33"},"PeriodicalIF":0.0,"publicationDate":"2022-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46111223","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mert Toslali, E. Ates, Darby Huye, Alex Ellis, Zhao Zhang, Lan Liu, Samantha Puterman, A. Coskun, Raja R. Sambasivan
{"title":"VAIF: Variance-driven Automated Instrumentation Framework","authors":"Mert Toslali, E. Ates, Darby Huye, Alex Ellis, Zhao Zhang, Lan Liu, Samantha Puterman, A. Coskun, Raja R. Sambasivan","doi":"10.1145/3544497.3544504","DOIUrl":"https://doi.org/10.1145/3544497.3544504","url":null,"abstract":"Developers use logs to diagnose performance problems in distributed applications. But, it is difficult to know a priori where logs are needed and what information in them is needed to help diagnose problems that may occur in the future. We summarize our work on the Variance-driven Automated Instrumentation Framework (VAIF), which runs alongside distributed applications. In response to newly-observed performance problems, VAIF automatically searches the space of possible instrumentation choices to enable the logs needed to help diagnose them. To work, VAIF combines distributed tracing (an enhanced form of logging) with insights about how response-time variance can be decomposed on the criticalpath portions of requests’ traces.","PeriodicalId":38935,"journal":{"name":"Operating Systems Review (ACM)","volume":"56 1","pages":"42-50"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"64052326","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Yuting Jiang, Yifan Xiong, L. Qu, Cheng Luo, Chen Tian, Peng Cheng, Y. Xiong
{"title":"Moneo: Monitoring Fine-grained Metrics Nonintrusively in AI Infrastructure","authors":"Yuting Jiang, Yifan Xiong, L. Qu, Cheng Luo, Chen Tian, Peng Cheng, Y. Xiong","doi":"10.1145/3544497.3544501","DOIUrl":"https://doi.org/10.1145/3544497.3544501","url":null,"abstract":"Cloud-based AI infrastructure is becoming increasingly important, especially on large-scale distributed training. To improve its efficiency and serviceability, real-time monitoring of the infrastructure and workload profiling are proved to be the effective approach empirically. However, cloud environment poses great challenges as service providers cannot interfere with their tenants’ workloads or touch user data, thus previous instrumentation-based monitoring approach cannot be applied, nor does the workload trace collection. In this paper, we propose Moneo, a non-intrusive cloudfriendly monitoring system for AI infrastructure. Moneo is capable of intelligently collecting the key architecture-level metrics at finer granularity in real-time without instrumenting or tracing the workloads, which has been deployed in real production cloud, Azure. We analyze the results reported by Moneo for typical large-scale distributed AI workloads from real deployment. Results demonstrate that Moneo can effectively help service providers understand the real resource usage patterns of various AI workloads and real networking requirements, so as to get valuable findings help improve the efficiency of cloud infrastructure and optimize the software stack with the consideration of the characteristic resource usage requirements for different AI workloads. This is a revised version of the symposium paper [23] presented in IEEE ICC 2022 originally.","PeriodicalId":38935,"journal":{"name":"Operating Systems Review (ACM)","volume":"56 1","pages":"18-25"},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"64052256","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Huaicheng Li, Martin L. Putra, Ronald Shi, Xing Lin, G. Ganger, Haryadi S. Gunawi
{"title":"lODA","authors":"Huaicheng Li, Martin L. Putra, Ronald Shi, Xing Lin, G. Ganger, Haryadi S. Gunawi","doi":"10.1145/3477132.3483573","DOIUrl":"https://doi.org/10.1145/3477132.3483573","url":null,"abstract":"Predictable latency on flash storage is a long-pursuit goal, yet, unpredictability stays due to the unavoidable disturbance from many well-known SSD internal activities. To combat this issue, the recent NVMe IO Determinism (IOD) interface advocates host-level controls to SSD internal management tasks. While promising, challenges remain on how to exploit it for truly predictable performance. We present IODA, an I/O deterministic flash array design built on top of small but powerful extensions to the IOD interface for easy deployment. IODA exploits data redundancy in the context of IOD for a strong latency predictability contract. In IODA, SSDs are expected to quickly fail an I/O on purpose to allow predictable I/Os through proactive data reconstruction. In the case of concurrent internal operations, IODA introduces busy remaining time exposure and predictable-latency-window formulation to guarantee predictable data reconstructions. Overall, IODA only adds 5 new fields to the NVMe interface and a small modification in the flash firmware, while keeping most of the complexity in the host OS. Our evaluation shows that IODA improves the 95-99.99th latencies by up to 75x. IODA is also the nearest to the ideal, no disturbance case compared to 7 state-of-the-art preemption, suspension, GC coordination, partitioning, tiny-tail flash controller, prediction, and proactive approaches.","PeriodicalId":38935,"journal":{"name":"Operating Systems Review (ACM)","volume":"112 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73513389","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"dSpace","authors":"Silvery Fu, S. Ratnasamy","doi":"10.1145/3477132.3483559","DOIUrl":"https://doi.org/10.1145/3477132.3483559","url":null,"abstract":"We present dSpace, an open and modular programming framework that aims to simplify and accelerate the development of smart space applications. To achieve this, dSpace provides two key building blocks~digivices that implement device control and actuation and digidata that process IoT data to generate events and insights. In addition, dSpace introduces novel abstractions - mount, yield, and pipe - via which digivices and digidata can be composed into higher-level abstractions. We apply dSpace to home automation systems and show how developers can easily and flexibly leverage these abstractions to support a wide range of home automation scenarios. Finally, we show how the dSpace concepts can be realized using a microservices-based architecture and implement dSpace as a Kubernetes-compatible framework.","PeriodicalId":38935,"journal":{"name":"Operating Systems Review (ACM)","volume":"55 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77992865","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Caracal","authors":"Dai Qin, Angela Demke Brown, Ashvin Goel","doi":"10.1145/3477132.3483591","DOIUrl":"https://doi.org/10.1145/3477132.3483591","url":null,"abstract":"Deterministic databases offer several benefits: they ensure serializable execution while avoiding concurrency-control related aborts, and they scale well in distributed environments. Today, most deterministic database designs use partitioning to scale up and avoid contention. However, partitioning requires significant programmer effort, leads to poor performance under skewed workloads, and incurs unnecessary overheads in certain uncontended workloads. We present the design of Caracal, a novel shared-memory, deterministic database that performs well under both skew and contention. Our deterministic scheme batches transactions in epochs and executes the transactions in an epoch in a predetermined order. Our scheme enables reducing contention by batching concurrency control operations. It also allows analyzing the transactions in the epoch to determine contended keys accurately. Certain transactions can then be split into independent contended and uncontended pieces and run deterministically and in parallel, further reducing contention. Based on these ideas, we present two novel optimizations, batch append and split-on-demand, for managing contention. With these optimizations, Caracal scales well and outperforms existing deterministic schemes in most workloads by 1.9x to 9.7x.","PeriodicalId":38935,"journal":{"name":"Operating Systems Review (ACM)","volume":"53 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81406144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sangmin Lee, Zhenhua Guo, Omer Sunercan, Jun Ying, Thawan Kooburat, Suryadeep Biswal, Jun Chen, K. Huang, Yatpang Cheung, Yiding Zhou, K. Veeraraghavan, Biren Damani, Pol Mauri Ruiz, V. Mehta, Chunqiang Tang
{"title":"Shard Manager: A Generic Shard Management Framework for Geo-distributed Applications","authors":"Sangmin Lee, Zhenhua Guo, Omer Sunercan, Jun Ying, Thawan Kooburat, Suryadeep Biswal, Jun Chen, K. Huang, Yatpang Cheung, Yiding Zhou, K. Veeraraghavan, Biren Damani, Pol Mauri Ruiz, V. Mehta, Chunqiang Tang","doi":"10.1145/3477132.3483546","DOIUrl":"https://doi.org/10.1145/3477132.3483546","url":null,"abstract":"Sharding is widely used to scale an application. Despite a decade of effort to build generic sharding frameworks that can be reused across different applications, the extent of their success remains unclear. We attempt to answer a fundamental question: what barriers prevent a sharding framework from getting adopted by the majority of sharded applications? We analyze hundreds of sharded applications at Facebook and identify two major barriers: 1) lack of support for geo-distributed applications, which account for most of Facebook's applications, and 2) inability to maintain application availability during planned events such as software upgrades, which happen ≈1000 times more frequently than unplanned failures. A sharding framework that does not help applications to address these fundamental challenges is not sufficiently attractive for most applications to adopt it. Other adoption barriers include the burden of supporting many complex applications in a one-size-fit-all sharding framework and the difficulty in supporting sophisticated shard-placement requirements. Theoretically, a constraint solver can handle complex placement requirements, but in practice it is not scalable enough to perform near-realtime shard placement at a global scale. We have overcome these adoption barriers in Facebook's sharding framework called Shard Manager. Currently, Shard Manager is used by hundreds of applications running on over one million machines, which account for about 54% of all sharded applications at Facebook.","PeriodicalId":38935,"journal":{"name":"Operating Systems Review (ACM)","volume":"25 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87152392","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Cuckoo Trie: Exploiting Memory-Level Parallelism for Efficient DRAM Indexing","authors":"Adar Zeitak, Adam Morrison","doi":"10.1145/3477132.3483551","DOIUrl":"https://doi.org/10.1145/3477132.3483551","url":null,"abstract":"We present the Cuckoo Trie, a fast, memory-efficient ordered index structure. The Cuckoo Trie is designed to have memory-level parallelism---which a modern out-of-order processor can exploit to execute DRAM accesses in parallel--- without sacrificing memory efficiency. The Cuckoo Trie thus breaks a fundamental performance barrier faced by current indexes, whose bottleneck is a series of dependent pointer-chasing DRAM accesses---e.g., traversing a search tree path--- which the processor cannot parallelize. Our evaluation shows that the Cuckoo Trie outperforms state-of-the-art-indexes by up to 20%-360% on a variety of datasets and workloads, typically with a smaller or comparable memory footprint.","PeriodicalId":38935,"journal":{"name":"Operating Systems Review (ACM)","volume":"57 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90339222","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
