Jaehong Min, Chenxingyu Zhao, Ming Liu, Arvind Krishnamurthy
{"title":"eZNS:增强性能隔离和设备利用率的弹性分区命名空间","authors":"Jaehong Min, Chenxingyu Zhao, Ming Liu, Arvind Krishnamurthy","doi":"10.1145/3653716","DOIUrl":null,"url":null,"abstract":"<p>Emerging Zoned Namespace (ZNS) SSDs, providing the coarse-grained zone abstraction, hold the potential to significantly enhance the cost-efficiency of future storage infrastructure and mitigate performance unpredictability. However, existing ZNS SSDs have a static zoned interface, making them in-adaptable to workload runtime behavior, unscalable to underlying hardware capabilities, and interfering with co-located zones. Applications either under-provision the zone resources yielding unsatisfied throughput, create over-provisioned zones and incur costs, or experience unexpected I/O latencies. </p><p>We propose eZNS, an elastic-zoned namespace interface that exposes an adaptive zone with predictable characteristics. eZNS comprises two major components: a zone arbiter that manages zone allocation and active resources on the control plane, a hierarchical I/O scheduler with read congestion control, and write admission control on the data plane. Together, eZNS enables the transparent use of a ZNS SSD and closes the gap between application requirements and zone interface properties. Our evaluations over RocksDB demonstrate that eZNS outperforms a static zoned interface by 17.7% and 80.3% in throughput and tail latency, respectively, at most.</p>","PeriodicalId":49113,"journal":{"name":"ACM Transactions on Storage","volume":"52 1","pages":""},"PeriodicalIF":2.1000,"publicationDate":"2024-04-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"eZNS: Elastic Zoned Namespace for Enhanced Performance Isolation and Device Utilization\",\"authors\":\"Jaehong Min, Chenxingyu Zhao, Ming Liu, Arvind Krishnamurthy\",\"doi\":\"10.1145/3653716\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>Emerging Zoned Namespace (ZNS) SSDs, providing the coarse-grained zone abstraction, hold the potential to significantly enhance the cost-efficiency of future storage infrastructure and mitigate performance unpredictability. However, existing ZNS SSDs have a static zoned interface, making them in-adaptable to workload runtime behavior, unscalable to underlying hardware capabilities, and interfering with co-located zones. Applications either under-provision the zone resources yielding unsatisfied throughput, create over-provisioned zones and incur costs, or experience unexpected I/O latencies. </p><p>We propose eZNS, an elastic-zoned namespace interface that exposes an adaptive zone with predictable characteristics. eZNS comprises two major components: a zone arbiter that manages zone allocation and active resources on the control plane, a hierarchical I/O scheduler with read congestion control, and write admission control on the data plane. Together, eZNS enables the transparent use of a ZNS SSD and closes the gap between application requirements and zone interface properties. Our evaluations over RocksDB demonstrate that eZNS outperforms a static zoned interface by 17.7% and 80.3% in throughput and tail latency, respectively, at most.</p>\",\"PeriodicalId\":49113,\"journal\":{\"name\":\"ACM Transactions on Storage\",\"volume\":\"52 1\",\"pages\":\"\"},\"PeriodicalIF\":2.1000,\"publicationDate\":\"2024-04-12\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"ACM Transactions on Storage\",\"FirstCategoryId\":\"94\",\"ListUrlMain\":\"https://doi.org/10.1145/3653716\",\"RegionNum\":3,\"RegionCategory\":\"计算机科学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"COMPUTER SCIENCE, HARDWARE & ARCHITECTURE\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"ACM Transactions on Storage","FirstCategoryId":"94","ListUrlMain":"https://doi.org/10.1145/3653716","RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"COMPUTER SCIENCE, HARDWARE & ARCHITECTURE","Score":null,"Total":0}
eZNS: Elastic Zoned Namespace for Enhanced Performance Isolation and Device Utilization
Emerging Zoned Namespace (ZNS) SSDs, providing the coarse-grained zone abstraction, hold the potential to significantly enhance the cost-efficiency of future storage infrastructure and mitigate performance unpredictability. However, existing ZNS SSDs have a static zoned interface, making them in-adaptable to workload runtime behavior, unscalable to underlying hardware capabilities, and interfering with co-located zones. Applications either under-provision the zone resources yielding unsatisfied throughput, create over-provisioned zones and incur costs, or experience unexpected I/O latencies.
We propose eZNS, an elastic-zoned namespace interface that exposes an adaptive zone with predictable characteristics. eZNS comprises two major components: a zone arbiter that manages zone allocation and active resources on the control plane, a hierarchical I/O scheduler with read congestion control, and write admission control on the data plane. Together, eZNS enables the transparent use of a ZNS SSD and closes the gap between application requirements and zone interface properties. Our evaluations over RocksDB demonstrate that eZNS outperforms a static zoned interface by 17.7% and 80.3% in throughput and tail latency, respectively, at most.
期刊介绍:
The ACM Transactions on Storage (TOS) is a new journal with an intent to publish original archival papers in the area of storage and closely related disciplines. Articles that appear in TOS will tend either to present new techniques and concepts or to report novel experiences and experiments with practical systems. Storage is a broad and multidisciplinary area that comprises of network protocols, resource management, data backup, replication, recovery, devices, security, and theory of data coding, densities, and low-power. Potential synergies among these fields are expected to open up new research directions.