{"title":"Zero-copy receive path in virtio","authors":"K. Meth, Mike Rapoport, J. Nider, Razya Ladelsky","doi":"10.1145/3078468.3078488","DOIUrl":"https://doi.org/10.1145/3078468.3078488","url":null,"abstract":"In the KVM hypervisor, incoming packets from the network must pass through several objects in the Linux kernel before being delivered to the guest VM. Currently, both the hypervisor and the guest keep their own sets of buffers on the receive path. For large packets, the overall processing time is dominated by the copying of data from hypervisor buffers to guest buffers.","PeriodicalId":249581,"journal":{"name":"Proceedings of the 10th ACM International Systems and Storage Conference","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129768572","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":"QoS differentiation in an advanced object store","authors":"Doron Chen, Gürkan Gür","doi":"10.1145/3078468.3081874","DOIUrl":"https://doi.org/10.1145/3078468.3081874","url":null,"abstract":"We present QoS experiments done in an advanced object store that supports the local execution of code on objects. We show how by giving different levels of resources to different objects, we are able to offer considerable improvement in the performance of the more-important objects, at the expense of the less-important objects. We created two tiers: Gold for premium objects, and Silver for standard objects. We present a workload on which we achieve a running time improvement of 33.5 percent for Gold objects. This is done at the expense of the performance given to Silver objects. The resource parameters we manipulate are the level of redundancy and CPU resources.","PeriodicalId":249581,"journal":{"name":"Proceedings of the 10th ACM International Systems and Storage Conference","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123855903","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}
Jingpei Yang, R. Pandurangan, Changho Choi, V. Balakrishnan
{"title":"AutoStream: automatic stream management for multi-streamed SSDs","authors":"Jingpei Yang, R. Pandurangan, Changho Choi, V. Balakrishnan","doi":"10.1145/3078468.3078469","DOIUrl":"https://doi.org/10.1145/3078468.3078469","url":null,"abstract":"Multi-stream SSDs can isolate data with different life time to disparate erase blocks, thus reduce garbage collection overhead and improve overall SSD performance. Applications are responsible for management of these device-level steams such as stream open/close and data-to-stream mapping. This requires application changes, and the engineer deploying the solution needs to be able to individually identify the streams in their workload. Furthermore, when multiple applications are involved, such as in VM or containerized environments, stream management becomes more complex due to the limited number of streams a device can support, for example, allocating streams to applications or sharing streams across applications will cause additional overhead. To address these issues and reduce the overhead of stream management, this paper proposes automatic stream management algorithms that operate under the application layer. Our stream assignment techniques, called AutoStream, is based on run time workload detection and independent of the application(s). We implement our AutoStream prototype in NVMe Linux device driver and our performance evaluation shows up to 60% reduction on WAF(Write Amplification Factor) and up to 237% improvement on performance compared to a conventional SSD device.","PeriodicalId":249581,"journal":{"name":"Proceedings of the 10th ACM International Systems and Storage Conference","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121431440","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}
G. Vernik, M. Factor, E. K. Kolodner, P. Michiardi, Effi Ofer, Francesco Pace
{"title":"Stocator: a high performance object store connector for spark","authors":"G. Vernik, M. Factor, E. K. Kolodner, P. Michiardi, Effi Ofer, Francesco Pace","doi":"10.1145/3078468.3078496","DOIUrl":"https://doi.org/10.1145/3078468.3078496","url":null,"abstract":"Data is the natural resource of the 21st century. It is being produced at dizzying rates, e.g., for genomics by sequencers, for Media and Entertainment with very high resolution formats, and for Internet of Things (IoT) by multitudes of sensors. Object Stores such as AWS S3, Azure Blob storage, and IBM Cloud Object Storage, are highly scalable distributed storage systems that offer high capacity, cost effective storage for this data. But it is not enough just to store data; we also need to derive value from it. Apache Spark is the leading big data analytics processing engine. It runs up to one hundred times faster than Hadoop MapReduce and combines SQL, streaming and complex analytics. In this poster we present Stocator, a high performance storage connector, that enables Spark to work directly on data stored in object storage systems.","PeriodicalId":249581,"journal":{"name":"Proceedings of the 10th ACM International Systems and Storage Conference","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126650830","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":"Remote page faults with a CAPI based FPGA","authors":"J. Nider, Yiftach Binyamini, Mike Rapoport","doi":"10.1145/3078468.3078489","DOIUrl":"https://doi.org/10.1145/3078468.3078489","url":null,"abstract":"Post-copy VM or container migration requires that the bulk of the memory is transferred after resuming on the destination node [3]. Transferring memory between nodes over a commodity TCP/IP network incurs too much latency, which slows down execution of the application on the destination node.","PeriodicalId":249581,"journal":{"name":"Proceedings of the 10th ACM International Systems and Storage Conference","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126517464","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":"Models for evaluating throughput","authors":"Netanel Zakay, D. Feitelson","doi":"10.1145/3078468.3081876","DOIUrl":"https://doi.org/10.1145/3078468.3081876","url":null,"abstract":"Analysts are interested in two categories of performance metrics: those concerned with time (response or wait time), and those concerned with rates (throughput or utilization, which reflect productivity and how well resources are used). In principle, these two categories are independent of each other, and both should be evaluated. But a common mistake is to \"measure\" the throughput in open-system evaluations, where the throughput is actually dictated directly by the workload. In order to evaluate throughput, the system model must include a feedback loop which modulates the workload being processed. The common solution is to create a pure closed system with a fixed number of users, who submit jobs in a loop. However, such behavior is often unrealistic. We review and analyze two alternative models that provide the required feedback by combining open and closed components: the mixed model which includes two such job classes, and the re-open model in which open user arrivals are combined with performance-dependent closed repetition of jobs by these users. These models allow evaluations of the trade-off between response time and throughput, including the throughput as it is observed by each user.","PeriodicalId":249581,"journal":{"name":"Proceedings of the 10th ACM International Systems and Storage Conference","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125144355","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}
Jens Kehne, Marius Hillenbrand, Jonathan Metter, Mathias Gottschlag, Martin Merkel, Frank Bellosa
{"title":"GPrioSwap: towards a swapping policy for GPUs","authors":"Jens Kehne, Marius Hillenbrand, Jonathan Metter, Mathias Gottschlag, Martin Merkel, Frank Bellosa","doi":"10.1145/3078468.3078474","DOIUrl":"https://doi.org/10.1145/3078468.3078474","url":null,"abstract":"Over the last few years, Graphics Processing Units (GPUs) have become popular in computing, and have found their way into a number of cloud platforms. However, integrating a GPU into a cloud environment requires the cloud provider to efficiently virtualize the GPU. While several research projects have addressed this challenge in the past, few of these projects attempt to properly enable sharing of GPU memory between multiple clients: To date, GPUswap is the only project that enables sharing of GPU memory without inducing unnecessary application overhead, while maintaining both fairness and high utilization of GPU memory. However, GPUswap includes only a rudimentary swapping policy, and therefore induces a rather large application overhead. In this paper, we work towards a practicable swapping policy for GPUs. To that end, we analyze the behavior of various GPU applications to determine their memory access patterns. Based on our insights about these patterns, we derive a swapping policy that includes a developer-assigned priority for each GPU buffer in its swapping decisions. Experiments with our prototype implementation show that a swapping policy based on buffer priorities can significantly reduce the swapping overhead.","PeriodicalId":249581,"journal":{"name":"Proceedings of the 10th ACM International Systems and Storage Conference","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133835811","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}
D. Lorenz, E. Raichstein, K. Barabash, H. Kolodner, L. Schour, Shelly Garion
{"title":"CogNETive: insights and visualization for operations@scale","authors":"D. Lorenz, E. Raichstein, K. Barabash, H. Kolodner, L. Schour, Shelly Garion","doi":"10.1145/3078468.3078495","DOIUrl":"https://doi.org/10.1145/3078468.3078495","url":null,"abstract":"Operating a cloud-scale service is a huge challenge. There are millions of users worldwide and millions of requests per seconds. For example, Amazon's Simple Storage Service (S3) in 2013 contained two trillion objects and its logs contained 1.1 million log lines per second, which are approximately 10 PB of log records per year (see [1]). Cloud scale implies thousands of servers and network elements, and hundreds of services from multiple cross-regional data centers. Cloud service operation data is scattered over various types of semi-structured and unstructured logs (e.g., application, error, debug), telemetry and network data, as well as customer service records. It is therefore extremely difficult for the multiple owners and administrators in such systems, coming from different units of the organization, to follow the possible paths and system alternatives in order to detect problems, solve issues and understand the service operation.","PeriodicalId":249581,"journal":{"name":"Proceedings of the 10th ACM International Systems and Storage Conference","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126799684","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":"Heterogeneous- and NUMA-aware scheduling for many-core architectures","authors":"P. Petrides, P. Trancoso","doi":"10.1145/3078468.3078482","DOIUrl":"https://doi.org/10.1145/3078468.3078482","url":null,"abstract":"As the number of cores increases in a single chip processor, several challenges arise: wire delays, contention for out-of-chip accesses, and core heterogeneity. In order to address these issues and the applications demands, future large-scale many-core processors are expected to be organized as a collection of NUMA clusters of heterogeneous cores. In this work we propose a scheduler that takes into account the non-uniform memory latency, the heterogeneity of the cores, and the contention to the memory controller to find the best matching core for the application's memory and compute requirements. Scheduler decisions are based on an on-line classification process that determines applications requirements either as memory- or compute-bound. We evaluate our proposed scheduler on the 48-core Intel SCC using applications from SPEC CPU2006 benchmark suite. Our results show that even when all cores are busy, migrating processes to cores that match better the requirements of applications results in overall performance improvement. In particular we observed a reduction of the execution time from 15% to 36% compared to a random static scheduling policy.","PeriodicalId":249581,"journal":{"name":"Proceedings of the 10th ACM International Systems and Storage Conference","volume":"102 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122504410","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":"Relieving self-healing SSDs of heal storms","authors":"Li-Pin Chang, Sheng-Min Huang, Kun-Lin Chou","doi":"10.1145/3078468.3078476","DOIUrl":"https://doi.org/10.1145/3078468.3078476","url":null,"abstract":"Building self-healing SSDs is proven feasible by recent studies. When the stress of a block becomes critical, it can be healed to remove part of the stress. However, with wear leveling, all blocks are evenly worn and have similar stress, and all blocks could undergo the healing process within a short period of time. The intensive heal operations, called heal storms, cause highly unpredictable I/O performance and storage reliability. Inspired by the even distribution of erase counts under wear leveling, we propose to operate wear leveling on virtual erase counts instead of real erase counts. When the balance among virtual erase counts is achieved through wear leveling, all real erase counts become evenly dispersed in a controlled interval. In this way, blocks will undergo healing at different times. Virtual erase counts are progressively adjusted such that all blocks reach their endurance limit when the SSD permanently retires. Our results show that our approach successfully resolved the heal storm problem without impacting on the SSD lifespan.","PeriodicalId":249581,"journal":{"name":"Proceedings of the 10th ACM International Systems and Storage Conference","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-05-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125711307","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}