Endre søk
Link to record
Permanent link

Direct link
BETA
Maguire Jr., Gerald Q., professorORCID iD iconorcid.org/0000-0002-6066-746X
Alternativa namn
Publikasjoner (10 av 302) Visa alla publikasjoner
Farshin, A., Roozbeh, A., Maguire Jr., G. Q. & Kostic, D. (2019). Make the Most out of Last Level Cache in Intel Processors. In: Proceedings of the Fourteenth EuroSys Conference (EuroSys'19), Dresden, Germany, 25-28 March 2019.: . Paper presented at EuroSys'19. ACM Digital Library
Åpne denne publikasjonen i ny fane eller vindu >>Make the Most out of Last Level Cache in Intel Processors
2019 (engelsk)Inngår i: Proceedings of the Fourteenth EuroSys Conference (EuroSys'19), Dresden, Germany, 25-28 March 2019., ACM Digital Library, 2019Konferansepaper, Publicerat paper (Fagfellevurdert)
Abstract [en]

In modern (Intel) processors, Last Level Cache (LLC) is divided into multiple slices and an undocumented hashing algorithm (aka Complex Addressing) maps different parts of memory address space among these slices to increase the effective memory bandwidth. After a careful study of Intel’s Complex Addressing, we introduce a slice-aware memory management scheme, wherein frequently used data can be accessed faster via the LLC. Using our proposed scheme, we show that a key-value store can potentially improve its average performance ∼12.2% and ∼11.4% for 100% & 95% GET workloads, respectively. Furthermore, we propose CacheDirector, a network I/O solution which extends Direct Data I/O (DDIO) and places the packet’s header in the slice of the LLC that is closest to the relevant processing core. We implemented CacheDirector as an extension to DPDK and evaluated our proposed solution for latency-critical applications in Network Function Virtualization (NFV) systems. Evaluation results show that CacheDirector makes packet processing faster by reducing tail latencies (90-99th percentiles) by up to 119 µs (∼21.5%) for optimized NFV service chains that are running at 100 Gbps. Finally, we analyze the effectiveness of slice-aware memory management to realize cache isolation

sted, utgiver, år, opplag, sider
ACM Digital Library, 2019
Emneord
Slice-aware Memory Management, Last Level Cache, Non-Uniform Cache Architecture, CacheDirector, DDIO, DPDK, Network Function Virtualization, Cache Partitioning, Cache Allocation Technology, Key-Value Store.
HSV kategori
Forskningsprogram
Datalogi
Identifikatorer
urn:nbn:se:kth:diva-244750 (URN)10.1145/3302424.3303977 (DOI)000470898700008 ()2-s2.0-85063919722 (Scopus ID)9781450362818 (ISBN)
Konferanse
EuroSys'19
Prosjekter
Time-Critical CloudsULTRAWASP
Forskningsfinansiär
Wallenberg AI, Autonomous Systems and Software Program (WASP)Swedish Foundation for Strategic Research EU, Horizon 2020, 770889
Merknad

QC 20190226

Tilgjengelig fra: 2019-02-24 Laget: 2019-02-24 Sist oppdatert: 2019-07-29bibliografisk kontrollert
Barbette, T., Katsikas, G. P., Maguire Jr., G. Q. & Kostic, D. (2019). RSS++: load and state-aware receive side scaling. In: ACM (Ed.), Proceedings of the 15th International Conference on emerging Networking EXperiments and Technologies: . Paper presented at CoNEXT' 19. Orlando, FL, USA: Association for Computing Machinery (ACM)
Åpne denne publikasjonen i ny fane eller vindu >>RSS++: load and state-aware receive side scaling
2019 (engelsk)Inngår i: Proceedings of the 15th International Conference on emerging Networking EXperiments and Technologies / [ed] ACM, Orlando, FL, USA: Association for Computing Machinery (ACM), 2019Konferansepaper, Publicerat paper (Fagfellevurdert)
Abstract [en]

While the current literature typically focuses on load-balancing among multiple servers, in this paper, we demonstrate the importance of load-balancing within a single machine (potentially with hundreds of CPU cores). In this context, we propose a new load-balancing technique (RSS++) that dynamically modifies the receive side scaling (RSS) indirection table to spread the load across the CPU cores in a more optimal way. RSS++ incurs up to 14x lower 95th percentile tail latency and orders of magnitude fewer packet drops compared to RSS under high CPU utilization. RSS++ allows higher CPU utilization and dynamic scaling of the number of allocated CPU cores to accommodate the input load, while avoiding the typical 25% over-provisioning. RSS++ has been implemented for both (i) DPDK and (ii) the Linux kernel. Additionally, we implement a new state migration technique, which facilitates sharding and reduces contention between CPU cores accessing per-flow data. RSS++ keeps the flow-state by groups that can be migrated at once, leading to a 20% higher efficiency than a state of the art shared flow table.

sted, utgiver, år, opplag, sider
Orlando, FL, USA: Association for Computing Machinery (ACM), 2019
Emneord
networking, load-balancing, packet scheduling, high-speed networking, intra-server load-balancing, receive side scaling, network function virtualization, RSS++
HSV kategori
Forskningsprogram
Informations- och kommunikationsteknik; Datalogi
Identifikatorer
urn:nbn:se:kth:diva-263941 (URN)10.1145/3359989.3365412 (DOI)978-1-4503-6998-5 (ISBN)
Konferanse
CoNEXT' 19
Forskningsfinansiär
Swedish Foundation for Strategic Research , TCCEU, European Research Council, 770889
Merknad

QC 20191126

Tilgjengelig fra: 2019-11-20 Laget: 2019-11-20 Sist oppdatert: 2019-11-26bibliografisk kontrollert
Bogdanov, K., Reda, W., Maguire Jr., G. Q., Kostic, D. & Canini, M. (2018). Fast and accurate load balancing for geo-distributed storage systems. In: SoCC 2018 - Proceedings of the 2018 ACM Symposium on Cloud Computing: . Paper presented at 2018 ACM Symposium on Cloud Computing, SoCC 2018, Carlsbad, United States, 11 October 2018 through 13 October 2018 (pp. 386-400). Association for Computing Machinery (ACM)
Åpne denne publikasjonen i ny fane eller vindu >>Fast and accurate load balancing for geo-distributed storage systems
Vise andre…
2018 (engelsk)Inngår i: SoCC 2018 - Proceedings of the 2018 ACM Symposium on Cloud Computing, Association for Computing Machinery (ACM), 2018, s. 386-400Konferansepaper, Publicerat paper (Fagfellevurdert)
Abstract [en]

The increasing density of globally distributed datacenters reduces the network latency between neighboring datacenters and allows replicated services deployed across neighboring locations to share workload when necessary, without violating strict Service Level Objectives (SLOs). We present Kurma, a practical implementation of a fast and accurate load balancer for geo-distributed storage systems. At run-time, Kurma integrates network latency and service time distributions to accurately estimate the rate of SLO violations for requests redirected across geo-distributed datacenters. Using these estimates, Kurma solves a decentralized rate-based performance model enabling fast load balancing (in the order of seconds) while taming global SLO violations. We integrate Kurma with Cassandra, a popular storage system. Using real-world traces along with a geo-distributed deployment across Amazon EC2, we demonstrate Kurma’s ability to effectively share load among datacenters while reducing SLO violations by up to a factor of 3 in high load settings or reducing the cost of running the service by up to 17%.

sted, utgiver, år, opplag, sider
Association for Computing Machinery (ACM), 2018
Emneord
Cloud Computing, Distributed Systems, Server Load Balancing, Service Level Objectives, Wide Area Networks
HSV kategori
Identifikatorer
urn:nbn:se:kth:diva-241481 (URN)10.1145/3267809.3267820 (DOI)000458692200031 ()2-s2.0-85059006718 (Scopus ID)9781450360111 (ISBN)
Konferanse
2018 ACM Symposium on Cloud Computing, SoCC 2018, Carlsbad, United States, 11 October 2018 through 13 October 2018
Forskningsfinansiär
EU, Horizon 2020, 770889Swedish Foundation for Strategic Research
Merknad

QC 20190123

Tilgjengelig fra: 2019-01-23 Laget: 2019-01-23 Sist oppdatert: 2019-10-17bibliografisk kontrollert
Bogdanov, K., Reda, W., Kostic, D., Maguire Jr., G. Q. & Canini, M. (2018). Kurma: Fast and Efficient Load Balancing for Geo-Distributed Storage Systems: Evaluation of Convergence and Scalability.
Åpne denne publikasjonen i ny fane eller vindu >>Kurma: Fast and Efficient Load Balancing for Geo-Distributed Storage Systems: Evaluation of Convergence and Scalability
Vise andre…
2018 (engelsk)Rapport (Annet vitenskapelig)
Abstract [en]

This report provides an extended evaluation of Kurma, a practical implementation of a geo-distributed load balancer for backend storage systems. In this report we demonstrate the ability of distributed Kurma instances to accurately converge to the same solutions within 1% of the total datacenter’s capacity and the ability of Kurma to scale up to 8 datacenters using a single CPU core at each datacenter.

HSV kategori
Identifikatorer
urn:nbn:se:kth:diva-222289 (URN)
Merknad

QR 20180212

Tilgjengelig fra: 2018-02-05 Laget: 2018-02-05 Sist oppdatert: 2018-02-12bibliografisk kontrollert
Roozbeh, A., Soares, J., Maguire Jr., G. Q., Wuhib, F., Padala, C., Mahloo, M., . . . Kostic, D. (2018). Software-Defined "Hardware" Infrastructures: A Survey on Enabling Technologies and Open Research Directions. IEEE Communications Surveys and Tutorials, 20(3), 2454-2485
Åpne denne publikasjonen i ny fane eller vindu >>Software-Defined "Hardware" Infrastructures: A Survey on Enabling Technologies and Open Research Directions
Vise andre…
2018 (engelsk)Inngår i: IEEE Communications Surveys and Tutorials, ISSN 1553-877X, E-ISSN 1553-877X, Vol. 20, nr 3, s. 2454-2485Artikkel i tidsskrift (Fagfellevurdert) Published
Abstract [en]

This paper provides an overview of software-defined "hardware" infrastructures (SDHI). SDHI builds upon the concept of hardware (HW) resource disaggregation. HW resource disaggregation breaks today's physical server-oriented model where the use of a physical resource (e.g., processor or memory) is constrained to a physical server's chassis. SDHI extends the definition of of software-defined infrastructures (SDI) and brings greater modularity, flexibility, and extensibility to cloud infrastructures, thus allowing cloud operators to employ resources more efficiently and allowing applications not to be bounded by the physical infrastructure's layout. This paper aims to be an initial introduction to SDHI and its associated technological advancements. This paper starts with an overview of the cloud domain and puts into perspective some of the most prominent efforts in the area. Then, it presents a set of differentiating use-cases that SDHI enables. Next, we state the fundamentals behind SDI and SDHI, and elaborate why SDHI is of great interest today. Moreover, it provides an overview of the functional architecture of a cloud built on SDHI, exploring how the impact of this transformation goes far beyond the cloud infrastructure level in its impact on platforms, execution environments, and applications. Finally, an in-depth assessment is made of the technologies behind SDHI, the impact of these technologies, and the associated challenges and potential future directions of SDHI.

sted, utgiver, år, opplag, sider
Institute of Electrical and Electronics Engineers (IEEE), 2018
Emneord
CR-software-defined infrastructure, resource disaggregation, cloud infrastructure, rack-scale, hyperscale computing, disaggregated DC
HSV kategori
Identifikatorer
urn:nbn:se:kth:diva-235270 (URN)10.1109/COMST.2018.2834731 (DOI)000443030500033 ()2-s2.0-85046804138 (Scopus ID)
Forskningsfinansiär
Swedish Foundation for Strategic Research Knut and Alice Wallenberg Foundation
Merknad

QC 20180919

Tilgjengelig fra: 2018-09-19 Laget: 2018-09-19 Sist oppdatert: 2018-11-23bibliografisk kontrollert
Yalew, S. D., Maguire Jr., G. Q., Haridi, S. & Correia, M. (2017). Hail to the Thief: Protecting Data from Mobile Ransomware with ransomSafeDroid. In: Gkoulalasdivanis, A Correia, MP Avresky, DR (Ed.), 2017 IEEE 16th International Symposium on Network Computing and Applications, NCA 2017: . Paper presented at 16th IEEE International Symposium on Network Computing and Applications, NCA 2017, Cambridge, United States, 30 October 2017 through 1 November 2017 (pp. 351-358). Institute of Electrical and Electronics Engineers (IEEE), 2017
Åpne denne publikasjonen i ny fane eller vindu >>Hail to the Thief: Protecting Data from Mobile Ransomware with ransomSafeDroid
2017 (engelsk)Inngår i: 2017 IEEE 16th International Symposium on Network Computing and Applications, NCA 2017 / [ed] Gkoulalasdivanis, A Correia, MP Avresky, DR, Institute of Electrical and Electronics Engineers (IEEE), 2017, Vol. 2017, s. 351-358Konferansepaper, Publicerat paper (Fagfellevurdert)
Abstract [en]

The growing popularity of Android and the increasing amount of sensitive data stored in mobile devices have lead to the dissemination of Android ransomware. Ransomware is a class of malware that makes data inaccessible by blocking access to the device or, more frequently, by encrypting the data; to recover the data, the user has to pay a ransom to the attacker. A solution for this problem is to backup the data. Although backup tools are available for Android, these tools may be compromised or blocked by the ransomware itself. This paper presents the design and implementation of RANSOMSAFEDROID, a TrustZone based backup service for mobile devices. RANSOMSAFEDROID is protected from malware by leveraging the ARM TrustZone extension and running in the secure world. It does backup of files periodically to a secure local persistent partition and pushes these backups to external storage to protect them from ransomware. Initially, RANSOMSAFEDROID does a full backup of the device filesystem, then it does incremental backups that save the changes since the last backup. As a proof-of-concept, we implemented a RANSOMSAFEDROID prototype and provide a performance evaluation using an i.MX53 development board.

sted, utgiver, år, opplag, sider
Institute of Electrical and Electronics Engineers (IEEE), 2017
HSV kategori
Identifikatorer
urn:nbn:se:kth:diva-225237 (URN)10.1109/NCA.2017.8171377 (DOI)000426971900053 ()2-s2.0-85046532213 (Scopus ID)9781538614655 (ISBN)
Konferanse
16th IEEE International Symposium on Network Computing and Applications, NCA 2017, Cambridge, United States, 30 October 2017 through 1 November 2017
Merknad

QC 20180403

Tilgjengelig fra: 2018-04-03 Laget: 2018-04-03 Sist oppdatert: 2018-05-22bibliografisk kontrollert
Yalew, S. D., Mendonca, P., Maguire Jr., G. Q., Haridi, S. & Correia, M. (2017). TruApp: A TrustZone-based Authenticity Detection Service for Mobile Apps. In: 2017 IEEE 13TH INTERNATIONAL CONFERENCE ON WIRELESS AND MOBILE COMPUTING, NETWORKING AND COMMUNICATIONS (WIMOB): . Paper presented at 13th IEEE International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob), OCT 09-11, 2017, Rome, ITALY. IEEE
Åpne denne publikasjonen i ny fane eller vindu >>TruApp: A TrustZone-based Authenticity Detection Service for Mobile Apps
Vise andre…
2017 (engelsk)Inngår i: 2017 IEEE 13TH INTERNATIONAL CONFERENCE ON WIRELESS AND MOBILE COMPUTING, NETWORKING AND COMMUNICATIONS (WIMOB), IEEE , 2017Konferansepaper, Publicerat paper (Fagfellevurdert)
Abstract [en]

In less than a decade, mobile apps became an integral part of our lives. In several situations it is important to provide assurance that a mobile app is authentic, i.e., that it is indeed the app produced by a certain company. However, this is challenging, as such apps can be repackaged, the user malicious, or the app tampered with by an attacker. This paper presents the design of TRUAPP, a software authentication service that provides assurance of the authenticity and integrity of apps running on mobile devices. TRUAPP provides such assurance, even if the operating system is compromised, by leveraging the ARM TrustZone hardware security extension. TRUAPP uses a set of techniques (static watermarking, dynamic watermarking, and cryptographic hashes) to verify the integrity of the apps. The service was implemented in a hardware board that emulates a mobile device, which was used to do a thorough experimental evaluation of the service.

sted, utgiver, år, opplag, sider
IEEE, 2017
Serie
IEEE International Conference on Wireless and Mobile Computing Networking and Communications-WiMOB, ISSN 2160-4886
HSV kategori
Identifikatorer
urn:nbn:se:kth:diva-222218 (URN)10.1109/WiMOB.2017.8115820 (DOI)000419818000108 ()2-s2.0-85041407068 (Scopus ID)978-1-5386-3839-2 (ISBN)
Konferanse
13th IEEE International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob), OCT 09-11, 2017, Rome, ITALY
Merknad

QC 20180205

Tilgjengelig fra: 2018-02-05 Laget: 2018-02-05 Sist oppdatert: 2019-04-15bibliografisk kontrollert
Olivecrona, H., Maguire, G. Q. ., Noz, M. E., Zeleznik, M. P., Kesteris, U. & Weidenhielm, L. (2016). A CT method for following patients with both prosthetic replacement and implanted tantalum beads: preliminary analysis with a pelvic model and in seven patients. Journal of Orthopaedic Surgery and Research, 11, Article ID 27.
Åpne denne publikasjonen i ny fane eller vindu >>A CT method for following patients with both prosthetic replacement and implanted tantalum beads: preliminary analysis with a pelvic model and in seven patients
Vise andre…
2016 (engelsk)Inngår i: Journal of Orthopaedic Surgery and Research, ISSN 1749-799X, E-ISSN 1749-799X, Vol. 11, artikkel-id 27Artikkel i tidsskrift (Fagfellevurdert) Published
Abstract [en]

Background: Radiostereometric analysis (RSA) is often used for evaluating implanted devices over time. Following patients who have had tantalum beads implanted as markers in conjunction with joint replacements is important for longitudinal evaluation of these patients and for those with similar implants. As doing traditional RSA imaging is exacting and limited to specialized centers, it is important to consider alternative techniques for this ongoing evaluation. This paper studies the use of computed tomography (CT) to evaluate over time tantalum beads which have been implanted as markers. Methods: The project uses both a hip model implanted with tantalum beads, acquired in several orientations, at two different CT energy levels, and a cohort of seven patients. The model was evaluated twice by the same observer with a 1-week interval. All CT volumes were analyzed using a semi-automated 3D volume fusion (spatial registration) tool which provides landmark-based fusion of two volumes, registering a target volume with a reference volume using a rigid body 3D algorithm. The mean registration errors as well as the accuracy and repeatability of the method were evaluated. Results: The mean registration error, maximum value of repeatability, and accuracy for the relative movement in the model were 0.16 mm, 0.02 degrees and 0.1 mm, and 0.36 degrees and 0.13 mm for 120 kVp and 0.21 mm, 0.04 degrees and 0.01 mm, and 0.39 degrees and 0.12 mm for 100 kVp. For the patients, the mean registration errors per patient ranged from 0.08 to 0.35 mm. These results are comparable to those in typical clinical RSA trials. This technique successfully evaluated two patients who would have been lost from the cohort if only RSA were used. Conclusions: The proposed technique can be used to evaluate patients with tantalum beads over time without the need for stereoradiographs. Further, the effective dose associated with CT is decreasing.

sted, utgiver, år, opplag, sider
BioMed Central, 2016
Emneord
Radiostereometry, Longitudinal studies, CT analysis, RSA
HSV kategori
Identifikatorer
urn:nbn:se:kth:diva-183622 (URN)10.1186/s13018-016-0360-7 (DOI)000370777100001 ()2-s2.0-84959129104 (Scopus ID)
Merknad

QC 20160319

Tilgjengelig fra: 2016-03-19 Laget: 2016-03-18 Sist oppdatert: 2017-11-30bibliografisk kontrollert
Brodén, C., Olivecrona, H., Maguire Jr., G. Q., Noz, M. E., Zeleznik, M. P. & Sköldenberg, O. (2016). Accuracy and Precision of Three-Dimensional Low Dose CT Compared to Standard RSA in Acetabular Cups: An Experimental Study. BioMed Research International, Article ID 5909741.
Åpne denne publikasjonen i ny fane eller vindu >>Accuracy and Precision of Three-Dimensional Low Dose CT Compared to Standard RSA in Acetabular Cups: An Experimental Study
Vise andre…
2016 (engelsk)Inngår i: BioMed Research International, ISSN 2314-6133, E-ISSN 2314-6141, artikkel-id 5909741Artikkel i tidsskrift (Fagfellevurdert) Published
Abstract [en]

Background and Purpose. The gold standard for detection of implant wear and migration is currently radiostereometry (RSA). The purpose of this study is to compare a three-dimensional computed tomography technique (3D CT) to standard RSA as an alternative technique for measuring migration of acetabular cups in total hip arthroplasty.

Materials and Methods. With tantalum beads, we marked one cemented and one uncemented cup and mounted these on a similarly marked pelvic model. A comparison was made between 3D CT and standard RSA for measuring migration. Twelve repeated stereoradiographs and CT scans with double examinations in each position and gradual migration of the implants were made. Precision and accuracy of the 3D CT were calculated.

Results. The accuracy of the 3D CT ranged between 0.07 and 0.32 mm for translations and 0.21 and 0.82° for rotation. The precision ranged between 0.01 and 0.09 mm for translations and 0.06 and 0.29° for rotations, respectively. For standard RSA, the precision ranged between 0.04 and 0.09 mm for translations and 0.08 and 0.32° for rotations, respectively. There was no significant difference in precision between 3D CT and standard RSA. The effective radiation dose of the 3D CT method, comparable to RSA, was estimated to be 0.33 mSv.

Interpretation. Low dose 3D CT is a comparable method to standard RSA in an experimental setting.

sted, utgiver, år, opplag, sider
Hindawi Publishing Corporation, 2016
HSV kategori
Identifikatorer
urn:nbn:se:kth:diva-189725 (URN)10.1155/2016/5909741 (DOI)000380310600001 ()2-s2.0-84979782542 (Scopus ID)
Merknad

QC 20190207

Tilgjengelig fra: 2016-07-11 Laget: 2016-07-11 Sist oppdatert: 2019-10-03bibliografisk kontrollert
Kim, Y. S., Maguire Jr., G. Q. & Noz, M. E. (2016). Do Small-Mass Neutrinos Participate in Gauge Transformations?. Advances in High Energy Physics, 1-7, Article ID 1847620.
Åpne denne publikasjonen i ny fane eller vindu >>Do Small-Mass Neutrinos Participate in Gauge Transformations?
2016 (engelsk)Inngår i: Advances in High Energy Physics, ISSN 1687-7357, E-ISSN 1687-7365, s. 1-7, artikkel-id 1847620Artikkel i tidsskrift (Fagfellevurdert) Published
Abstract [en]

Neutrino oscillation experiments presently suggest that neutrinos have a small but finite mass. If neutrinos have mass, there should be a Lorentz frame in which they can be brought to rest. This paper discusses how Wigner’s little groups can be used to distinguish between massive and massless particles. We derive a representation of the SL(2,c) group which separates out the two sets of spinors: one set is gauge dependent and the other set is gauge invariant and represents polarized neutrinos. We show that a similar calculation can be done for the Dirac equation. In the large-momentum/zero-mass limit, the Dirac spinors can be separated into large and small components. The large components are gauge invariant, while the small components are not. These small components represent spin-1/2 non-zero-mass particles. If we renormalize the large components, these gauge invariant spinors represent the polarization of neutrinos. Massive neutrinos cannot be invariant under gauge transformations.

sted, utgiver, år, opplag, sider
Hindawi Publishing Corporation, 2016
HSV kategori
Identifikatorer
urn:nbn:se:kth:diva-189756 (URN)10.1155/2016/1847620 (DOI)000385145700001 ()2-s2.0-84979656442 (Scopus ID)
Merknad

QC 20160713

Tilgjengelig fra: 2016-07-13 Laget: 2016-07-13 Sist oppdatert: 2019-10-28bibliografisk kontrollert
Organisasjoner
Identifikatorer
ORCID-id: ORCID iD iconorcid.org/0000-0002-6066-746X