Abstract
Petascale computing requires I/O subsystems that can keep up with the dramatic computing power demanded by such systems. TOP500.org ranks top computers based on their peak compute performance, but there has not been adequate investigation of the current state-of-the-art and future requirements of storage area networks that support petascale computers. Dependable scaling of an I/O subsystem to support petascale computing is not as simple as adding more storage servers. In this paper, we present a stochastic activity network model that uses failure rates computed from real logs to predict the reliability and availability of the storage architecture of the Abe cluster at the National Center for Supercomputing Applications (NCSA). We then use the model to evaluate the challenges encountered as one scales the number of storage servers to support petascale computing. The results present new insights regarding the dependability challenges that will be encountered when building next-generation petabyte storage. Furthermore, we provide insight into a new design approach that will enable system designers to integrate the trace-based analysis of parameter values from real system data into their stochastic models to allow informed design choices.
| Original language | English (US) |
|---|---|
| Pages | 386-391 |
| Number of pages | 6 |
| DOIs | |
| State | Published - Oct 13 2008 |
| Event | 2008 International Conference on Dependable Systems and Networks, DSN-2008 - Anchorage, AK, United States Duration: Jun 24 2008 → Jun 27 2008 |
Other
| Other | 2008 International Conference on Dependable Systems and Networks, DSN-2008 |
|---|---|
| Country | United States |
| City | Anchorage, AK |
| Period | 6/24/08 → 6/27/08 |
Fingerprint
Keywords
- Data analysis
- Modeling techniques
- Reliability and availability
- Simulation
- Storage systems
ASJC Scopus subject areas
- Software
- Hardware and Architecture
- Computer Networks and Communications
Cite this
Scaling file systems to support petascale clusters : A dependability analysis to support informed design choices. / Gaonkar, Shravan; Rozier, Eric; Tong, Anthony; Sanders, William H.
2008. 386-391 Paper presented at 2008 International Conference on Dependable Systems and Networks, DSN-2008, Anchorage, AK, United States.Research output: Contribution to conference › Paper
}
TY - CONF
T1 - Scaling file systems to support petascale clusters
T2 - A dependability analysis to support informed design choices
AU - Gaonkar, Shravan
AU - Rozier, Eric
AU - Tong, Anthony
AU - Sanders, William H.
PY - 2008/10/13
Y1 - 2008/10/13
N2 - Petascale computing requires I/O subsystems that can keep up with the dramatic computing power demanded by such systems. TOP500.org ranks top computers based on their peak compute performance, but there has not been adequate investigation of the current state-of-the-art and future requirements of storage area networks that support petascale computers. Dependable scaling of an I/O subsystem to support petascale computing is not as simple as adding more storage servers. In this paper, we present a stochastic activity network model that uses failure rates computed from real logs to predict the reliability and availability of the storage architecture of the Abe cluster at the National Center for Supercomputing Applications (NCSA). We then use the model to evaluate the challenges encountered as one scales the number of storage servers to support petascale computing. The results present new insights regarding the dependability challenges that will be encountered when building next-generation petabyte storage. Furthermore, we provide insight into a new design approach that will enable system designers to integrate the trace-based analysis of parameter values from real system data into their stochastic models to allow informed design choices.
AB - Petascale computing requires I/O subsystems that can keep up with the dramatic computing power demanded by such systems. TOP500.org ranks top computers based on their peak compute performance, but there has not been adequate investigation of the current state-of-the-art and future requirements of storage area networks that support petascale computers. Dependable scaling of an I/O subsystem to support petascale computing is not as simple as adding more storage servers. In this paper, we present a stochastic activity network model that uses failure rates computed from real logs to predict the reliability and availability of the storage architecture of the Abe cluster at the National Center for Supercomputing Applications (NCSA). We then use the model to evaluate the challenges encountered as one scales the number of storage servers to support petascale computing. The results present new insights regarding the dependability challenges that will be encountered when building next-generation petabyte storage. Furthermore, we provide insight into a new design approach that will enable system designers to integrate the trace-based analysis of parameter values from real system data into their stochastic models to allow informed design choices.
KW - Data analysis
KW - Modeling techniques
KW - Reliability and availability
KW - Simulation
KW - Storage systems
UR - http://www.scopus.com/inward/record.url?scp=53349175680&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=53349175680&partnerID=8YFLogxK
U2 - 10.1109/DSN.2008.4630107
DO - 10.1109/DSN.2008.4630107
M3 - Paper
AN - SCOPUS:53349175680
SP - 386
EP - 391
ER -