TY - GEN
T1 - SSD Failures in datacenters
T2 - 13th Joint International Conference on Measurement and Modeling of Computer Systems, ACM SIGMETRICS / IFIP Performance 2016
AU - Narayanan, Iyswarya
AU - Wang, Di
AU - Jeon, Myeongjae
AU - Sharma, Bikash
AU - Caulfield, Laura
AU - Sivasubramaniam, Anand
AU - Cutler, Ben
AU - Liu, Jie
AU - Khessib, Badriddine
AU - Vaid, Kushagra
N1 - Publisher Copyright:
© 2016 Copyright held by the owner/author(s). ACM.
PY - 2016/6/14
Y1 - 2016/6/14
N2 - Despite the growing popularity of Solid State Disks (SSDs) in the datacenter, little is known about their reliability characteristics in the field. The little knowledge is mainly vendor supplied, which cannot really help understand how SSD failures can manifest and impact production systems, in order to take appropriate actions. Besides failure data, a detailed characterization requires wide spectrum of data about factors influencing SSD failures, right from provisioning (what models? where and when deployed? etc.) to the operational ones (workloads, read-write intensities, write amplification, etc.). We analyze over half a million SSDs that span multiple generations spread across several datacenters which host a wide range of workloads over nearly 3 years. By studying the diverse set of factors on SSD failures, and their symptoms, our work provides the first look at the what, when and why characteristics of SSD failures in production datacenters.
AB - Despite the growing popularity of Solid State Disks (SSDs) in the datacenter, little is known about their reliability characteristics in the field. The little knowledge is mainly vendor supplied, which cannot really help understand how SSD failures can manifest and impact production systems, in order to take appropriate actions. Besides failure data, a detailed characterization requires wide spectrum of data about factors influencing SSD failures, right from provisioning (what models? where and when deployed? etc.) to the operational ones (workloads, read-write intensities, write amplification, etc.). We analyze over half a million SSDs that span multiple generations spread across several datacenters which host a wide range of workloads over nearly 3 years. By studying the diverse set of factors on SSD failures, and their symptoms, our work provides the first look at the what, when and why characteristics of SSD failures in production datacenters.
UR - http://www.scopus.com/inward/record.url?scp=84978766616&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84978766616&partnerID=8YFLogxK
U2 - 10.1145/2896377.2901489
DO - 10.1145/2896377.2901489
M3 - Conference contribution
AN - SCOPUS:84978766616
T3 - SIGMETRICS/ Performance 2016 - Proceedings of the SIGMETRICS/Performance Joint International Conference on Measurement and Modeling of Computer Science
SP - 407
EP - 408
BT - SIGMETRICS/ Performance 2016 - Proceedings of the SIGMETRICS/Performance Joint International Conference on Measurement and Modeling of Computer Science
PB - Association for Computing Machinery, Inc
Y2 - 14 June 2016 through 18 June 2016
ER -