Performance Issues and Optimizations for Block-level Network Storage M. Marazakis, V. Papaefstathiou and A. Bilas Networked Storage using Commodity Hardware Motivation:
Challenges for Commodity: • Systems software overheads • Increased complexity • Hardware-software orchestration • Performance gap Approach: • Minimal NIC architecture, 10 Gbps
– Primary networked storage subsystems – Consolidation of storage in one subsystem – Single interconnect for application and storage nodes
– Specialized controllers and interconnects – Expensive and difficult to scale – High incremental cost for increase in performance
• Block-level remote I/O protocol (kernel-space)
Objectives: • Understand the performance gap – By measurements on real system
• Address the associated problems – By applying targeted optimizations
Performance-limiting Factors:: • %CPU util at both Initiator & Target (#IRQs) • Inefficient use of NIC (# small RDMAs) • Initiator & Target execute either send-path or recv-path (exclusively / ctxsw via IRQ) • Disk & network transfers interfere at Target
Optimizations:: •IRQ batching at both Initiator & Target
500
•Asynchronous event processing in the send-path •Polling to capture “events” as soon as possible after completing normal sendpath processing
0
512
64
I/O Request Size (KB)
128
256
I/O Request Size (KB)
512
•Both at Initiator & Target
Summary of Results Quantify overheads at high network speeds & examine various protocol optimization techniques on a real system I/O Throughput: High-score Summary
References • “Efficient remote block-level I/O over an RDMA-capable NIC”, Proceedings of ACM ICS’06 • “Experiences from Debugging a PCIX-based RDMA-capable NIC”, Proceedings of RAIT’06 workshop (in conjunction with IEEE Cluster’06) • “Optimization and Bottleneck Analysis of Network Block I/O in Commodity Storage Systems”, Proceedings of ACM ICS’07
Foundation for Research and Technology – Hellas (FORTH) Institute of Computer Science (ICS)
such as routing, computer graphics, and networking [15], ...... PUC Minas in 2004 and his B.Sc. in Computer ... of Computer Science at the University of Edin-.
develop one hour of training., but we are now in an environ- ment where learning is ... in-person instructor-led training program, several hours for an. eLearning ...
semantics, we extend the data flow analysis framework to concurrent threads. .... duce two auxiliary conceptsâEvent Vector and Sensitivity Vectorâin section 6, ...
In this paper we study about social network theory and privacy challenges which affects a secure range of ... In recent years online social networking has moved from niche phenomenon to mass adoption. The rapid .... OSN users are leveraged by governm
In this paper we study about social network theory and privacy challenges which affects ... applications, such as recommender systems, email filtering, defending ...
prediction techniques provide fast lookup and power efficiency .... almost as fast as global prediction. .... efficient hash functions enable the better use of PHT.
pad memories (spms), which we call L0 instruction spms. ...... showing that direct-mapped filter caches outperform 4-way associative filter caches ...... 5When I presented a shorter version of this chapter at a conference, one of the most common.
Feb 1, 2011 - 1D standard was designated at a time where recovering network connectivity within about 60 seconds after an outage was considered as .... port receiving the best Bridge Protocol Data Unit (BPDU) on a bridge is a âroot portâ. ... as
the address computations from the data computations. De- coupled .... following SLIW âgroupingsâ (SLIW bundle templates):. Group .... definition sharing the same register. Another ..... part by the hardware loop mapping and the IF-conversion,.
Permission to make digital or hard copies of all or part of this work for personal or classroom use is .... In the SLIW mode, the data dependences are scoreboarded, provided they hold ...... servo Hard disk drive digital control loop. efr 5.1.0 ETSI
(Actually the expected value is a little more than $5 if we do not shuffle the pack after each pick and you are strategic). ⢠If the prize is doubled, you get two tries to ...
Any reduction in simulation time directly leads to productivity ...... In Proceedings of the 32nd Annual ACM/IEEE Design Automation Conference, 151â156.
DISCLAIMER AND LEGAL INFORMATION. All opinions expressed in this document are those of the authors individually and are not reflective or indicative of the opinions and positions of the authors' employers. The technology described in this document is
ââ¬ËCall Us Ms ââ¬â¢ Viva and arguments for Kenyan women s respectable ... access to affordable healthcare childcare and education reproductive BibMe Free ...
