Ceph Parallel File System Evaluation Report

Wang, Feiyi; Oral, Sarp; Fuller, Douglas; Caldwell, Blake; Simmons, James A.; Settlemyer, Bradley W.; Atchley, Scott; Hill, Jason; Nelson, Mark

doi:10.2172/1104795

Cited by 3 publications

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…SFA10K organizes disks into various RAID levels by two active-active RAID controllers. In our testbed, exported RAID groups (LUNs) are driven by four server hosts (oss- [1][2][3][4]). Each server host has two InfiniBand (IB) QDR connections (supported by a dual-port Mellanox ConnectX IB card) to SFA10K.…”

Section: Testbed Environment and Test Methodologymentioning

confidence: 99%

“…The result of this parameter probing is illustrated in Figure 6. Please refer to Appendix E of [4] for explanations of these probed parameters. In this figure, the default is where we started with.…”

Section: Further Optimizationmentioning

confidence: 99%

“…This behavior was first observed roughly in the kernel 3.5 time frame which was the kernel installed on our test systems. 4 We upgraded our test systems with kernel version 3.9 and performed RADOS Bench test. The results were dramatic: read performance improved from 4678 MB/s to 7499 MB/s (a 60% improvement).…”

Section: Improving Ior Read Performancementioning

confidence: 99%

See 2 more Smart Citations

Performance and scalability evaluation of the Ceph parallel file system

Wang

Nelson²,

Oral

et al. 2013

Proceedings of the 8th Parallel Data Storage Workshop

Self Cite

View full text Add to dashboard Cite

Ceph is an emerging open-source parallel distributed file and storage system. By design, Ceph leverages unreliable commodity storage and network hardware, and provides reliability and fault-tolerance via controlled object placement and data replication. This paper presents our file and block I/O performance and scalability evaluation of Ceph for scientific high-performance computing (HPC) environments. Our work makes two unique contributions. First, our evaluation is performed under a realistic setup for a large-scale capability HPC environment using a commercial high-end storage system. Second, our path of investigation, tuning efforts, and findings made direct contributions to Ceph's development and improved code quality, scalability, and performance. These changes should benefit both Ceph and the HPC community at large.

show abstract

Section: Testbed Environment and Test Methodologymentioning

confidence: 99%

Section: Further Optimizationmentioning

confidence: 99%