ZygOS: Achieving Low Tail Latency for Microsecond-scale Networked Tasks
Open accessThe DOI matches its Crossref record; 37 of the 37 references it cites were found in a registry or an index.
Abstract
This paper focuses on the efficient scheduling on multicore systems of very fine-grain networked tasks, which are the typical building block of online data-intensive applications. The explicit goal is to deliver high throughput (millions of remote procedure calls per second) for tail latency service-level objectives that are a small multiple of the task size.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Dean, J., & Barroso, L. A. (2013) · Communications of the ACM · doi:10.1145/2408776.2408794 · Cited by 1,852 · cited 2×
“Such applications must support high concurrent connection counts and operate with user-facing SLO, often defined in terms of tail latency to meet business objectives [1, 12, 47].”
BackgroundThe Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines
Barroso, L. A., & Hölzle, U. (2009) · Synthesis lectures on computer architecture · doi:10.2200/s00193ed1v01y200905cac006 · Cited by 1,714
kvm : the Linux Virtual Machine Monitor
Kivity, A. (2007) · Cited by 1,500
Dragojević, A., Narayanan, D., Hodson, O., & Castro, M. (2014) · Networked Systems Design and Implementation · doi:10.5555/2616448.2616486 · Cited by 1,016
Nishtala, R., Fugal, H., Grimm, S., Kwiatkowski, M., Lee, H., Li, H. C., McElroy, R., Paleczny, M., Peek, D., Saab, P., et al. (2013) · Networked Systems Design and Implementation · Cited by 978
Workload analysis of a large-scale key-value store
Atikoglu, B., Xu, Y., Frachtenberg, E., Jiang, S., & Paleczny, M. (2012) · Proceedings of the 12th ACM SIGMETRICS/PERFORMANCE joint international conference on Measurement and Modeling of Computer Systems · doi:10.1145/2254756.2254766 · Cited by 901 · cited 3×
“, some widely-studied memcached workloads [1, 43] with an SLO at > 100× the mean service time [5]), such assumptions break down when considering more complex workloads, e.”
BackgroundMethodIntel virtualization technology
Uhlig, R., Neiger, G., Rodgers, D., Santoni, A. L., Martins, F. C. M., Anderson, A. V., Bennett, S., Kägi, A., Leung, F., & Smith, L. M. (2005) · Computer · doi:10.1109/mc.2005.163 · Cited by 757
Speedy transactions in multicore in-memory databases
Tu, S., Zheng, W., Kohler, E., Liskov, B. H., & Madden, S. R. (2013) · ACM SIGOPS Symposium on Operating Systems Principles (SOSP) · doi:10.1145/2517349.2522713 · Cited by 688
Sparrow: distributed, low latency scheduling
Ousterhout, K., Wendell, P., Zaharia, M., & Stoica, I. (2013) · Proceedings of the Twenty-Fourth ACM Symposium on Operating Systems Principles · doi:10.1145/2517349.2522716 · Cited by 586 · cited 1×
“Sparrow [49] also relies on power-of-two choices for batch job scheduling.”
BackgroundThe end of an architectural era: (it's time for a complete rewrite)
Stonebraker, M. R., Madden, S. R., Abadi, D. J., Harizopoulos, S., Hachem, N. I., & Helland, P. (2007) · Cited by 558
Barroso, L. A., Clidaras, J., & Hölzle, U. (2013) · Synthesis lectures on computer architecture · doi:10.2200/s00516ed2v01y201306cac024 · Cited by 489 · cited 1×
“To meet service-level objectives (SLO), web-scale online dataintensive applications such as search, e-commerce, and social applications rely on the scale-out architectures of modern, warehouse-scale datacenters [2].”
BackgroundJoin-Idle-Queue: A novel load balancing algorithm for dynamically scalable web services
Lu, Y., Xie, Q., Kliot, G., Geller, A., Larus, J. R., & Greenberg, A. G. (2011) · Performance Evaluation · doi:10.1016/j.peva.2011.07.015 · Cited by 441
SAP HANA database: data management for modern business applications
Färber, F., Cha, S. K., Primsch, J., Bornhövd, C., Sigg, S., & Lehner, W. (2012) · ACM SIGMOD Record · doi:10.1145/2094114.2094126 · Cited by 396 · cited 1×
“, the social graph) in the memory of hundreds of data services, such as memory-resident transactional databases [20, 62, 65, 67, 68], NoSQL databases [44, 55], keyvalue stores [15, 39, 43, 74], or specialized graph stores [6].”
BackgroundMICA: A Holistic Approach to Fast In-Memory Key-Value Storage
Lim, H., Han, D., Andersen, D. G., & Kaminsky, M. (2018) · Figshare · doi:10.1184/r1/6607280 · Cited by 361
Fast in-memory transaction processing using RDMA and HTM
Wei, X., Shi, J., Chen, Y., Chen, R., & Chen, H. (2015) · ACM SIGOPS Symposium on Operating Systems Principles (SOSP) · doi:10.1145/2815400.2815419 · Cited by 327
mTCP: a highly scalable user-level TCP stack for multicore systems
Jeong, E. Y., Woo, S., Jamshed, M. A., Jeong, H., Ihm, S., Han, D., & Park, K. (2014) · doi:10.5555/2616448.2616493 · Cited by 271
Ousterhout, J., Gopalan, A., Gupta, A., Kejriwal, A., Lee, C., Montazeri, B., Ongaro, D., Park, S. J., Qin, H., Rosenblum, M., et al. (2015) · ACM Transactions on Computer Systems · doi:10.1145/2806887 · Cited by 268
The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines
Barroso, L. A., Clidaras, J., & Hölzle, U. (2013) · Synthesis Lectures on Computer Architecture · doi:10.1007/978-3-031-01741-4 · Cited by 259
Dynamic circular work-stealing deque
Chase, D., & Lev, Y. (2005) · Proceedings of the seventeenth annual ACM symposium on Parallelism in algorithms and architectures · doi:10.1145/1073970.1073974 · Cited by 238
Tales of the Tail: Hardware, OS, and Application-level Sources of Tail Latency
Li, J., Sharma, N. K., Ports, D. R. K., & Gribble, S. D. (2014) · Proceedings of the ACM Symposium on Cloud Computing · doi:10.1145/2670979.2670988 · Cited by 235
FlexSC: flexible system call scheduling with exception-less system calls
Soares, L., & Stumm, M. (2010) · doi:10.5555/1924943.1924946 · Cited by 223
Reconciling high server utilization and sub-millisecond quality-of-service
Leverich, J., & Kozyrakis, C. (2014) · Proceedings of the Ninth European Conference on Computer Systems · doi:10.1145/2592798.2592821 · Cited by 220 · cited 5×
“The client machines run mutilate [34] as a load generator: 10 machines generate load and the 11th one measures latency.”
BackgroundMethodInfluentialChronos: predictable low latency for data center applications
Kapoor, R., Porter, G., Tewari, M., Voelker, G. M., & Vahdat, A. M. (2012) · ACM Symposium on Cloud Computing (SoCC) · doi:10.1145/2391229.2391238 · Cited by 213 · cited 1×
“The latter improves throughput substantially (by up to 6× for key-value stores [5]) through sweeping simplifications such as separation of control from the dataplane execution, polling, run-to-completion, and synchronization-free, flow-consistent mapping of requests to cores [5, 26, 27, 39, 42, 51].”
BackgroundAttack of the killer microseconds
Barroso, L., Marty, M., Patterson, D., & Ranganathan, P. (2017) · Communications of the ACM · doi:10.1145/3015146 · Cited by 204
Leung, K.-C., Li, V. O. K., & Yang, D. (2007) · IEEE Transactions on Parallel and Distributed Systems · doi:10.1109/tpds.2007.1011 · Cited by 196 · cited 1×
“When TCP sockets are statically assigned to threads, applications can rely on intuitive ordering and concurrency semantics [33].”
BackgroundDune: safe user-level access to privileged CPU features
Belay, A., Bittau, A., Mashtizadeh, A. J., Terei, D. A., Mazières, D., & Kozyrakis, C. (2012) · doi:10.5555/2387880.2387913 · Cited by 190
ELI: bare-metal performance for I/O virtualization
Gordon, A., Amit, N., Har'El, N., Ben-Yehuda, M., Landau, A., Schuster, A., & Tsafrir, D. (2012) · Proceedings of the seventeenth international conference on Architectural Support for Programming Languages and Operating Systems · doi:10.1145/2150976.2151020 · Cited by 181 · cited 3×
“Using an approach similar to ELI [22], we added support in Dune for exit-less interrupts in non-root mode, based on the assumption that Z YG OS kernel’s interrupt handler will redirect to the Linux host operating system the interrupts that are destined to it.”
MethodCharacterizing and improving the performance of Intel Threading Building Blocks
Contreras, G., & Martonosi, M. (2008) · IEEE International Symposium on Workload Characterization · doi:10.1109/iiswc.2008.4636091 · Cited by 141
Network stack specialization for performance
Marinos, I., Watson, R. N. M., & Handley, M. J. (2014) · ACM SIGCOMM Conference (SIGCOMM) · doi:10.1145/2619239.2626311 · Cited by 126
Arrakis: The Operating System Is the Control Plane
Peter, S., Li, J., Zhang, I. Y., Ports, D. R. K., Woos, D., Krishnamurthy, A., Anderson, T. E., & Roscoe, T. (2015) · ACM Transactions on Computer Systems · doi:10.1145/2812806 · Cited by 116 · cited 4×
“The latter improves throughput substantially (by up to 6× for key-value stores [5]) through sweeping simplifications such as separation of control from the dataplane execution, polling, run-to-completion, and synchronization-free, flow-consistent mapping of requests to cores [5, 26, 27, 39, 42, 51].”
BackgroundMethodInfluentialEfficient and Available In-Memory KV-Store with Hybrid Erasure Coding and Replication
Chen, H., Zhang, H., Dong, M., Wang, Z., Xia, Y., Guan, H., & Zang, B. (2017) · ACM Transactions on Storage · doi:10.1145/3129900 · Cited by 94
Energy proportionality and workload consolidation for latency-critical applications
Prekas, G., Primorac, M., Belay, A., Kozyrakis, C., & Bugnion, E. (2015) · ACM Symposium on Cloud Computing (SoCC) · doi:10.1145/2806777.2806848 · Cited by 90 · cited 1×
“Control plane interactions: The IX control plane implement energy proportionality or workload consolidation by dynamically adjusting processor frequency and core allocation [54].”
MethodSelf-citationBelay, A., Prekas, G., Primorac, M., Klimovic, A., Grossman, S., Kozyrakis, C., & Bugnion, E. (2016) · ACM Transactions on Computer Systems · doi:10.1145/2997641 · Cited by 87
Self-citationFew-to-Many: Incremental Parallelism for Reducing Tail Latency in Interactive Services
Haque, M. E., Eom, Y. H., He, Y., Elnikety, S., Bianchini, R., & McKinley, K. S. (2015) · International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS) · doi:10.1145/2694344.2694384 · Cited by 86
The Scalable Commutativity Rule: Designing Scalable Software for Multicore Processors
Clements, A. T., Kaashoek, M. F., Zeldovich, N., Morris, R., & Kohler, E. (2015) · ACM Transactions on Computer Systems · doi:10.1145/2699681 · Cited by 78
Efficient and available in-memory KV-store with hybrid erasure coding and replication
Zhang, H., Dong, M., & Chen, H. (2016) · File and Storage Technologies · Cited by 75
Work stealing for interactive services to meet target latency
Li, J., Agrawal, K., Elnikety, S., He, Y., Lee, I.-T. A., Lu, C., & McKinley, K. S. (2016) · ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming · doi:10.1145/2851141.2851151 · Cited by 39
Metadata
- Authors
- Prekas, G., Kogias, M., & Bugnion, E.
- Venue
- Proceedings of the 26th Symposium on Operating Systems Principles
- Publisher
- ACM
- Published
- 2017-10-14
- Pages
- 325-341
- OpenAlex
- W2761028280
- Cited by
- 207
- References
- 37
- FWCI
- 10.08
- Type
- conference paper
- Language
- en
- Open access
- bronze
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.