Title :
A high-performance cluster storage server
Author :
Bell, Keith ; Chien, Andrew ; Lauria, Mario
Author_Institution :
Dept. of Comput. Sci. & Eng., California Univ., San Diego, CA, USA
Abstract :
An essential building block for any data grid infrastructure is the storage server. In this paper we describe a high-performance cluster storage server built around the SDSC Storage Resource Broker (SRB) and commodity workstations. A number of performance critical design issues and our solutions to them are described. We incorporate pipeline optimizations into SRB to enable the full overlapping of communication and disk I/O. With these optimizations we were able to deliver to the application more than 95% of the disk throughput achievable through a remote connection. Then we show how our approach to network-striped transport is effective in achieving aggregate cluster-to-cluster throughput which scales with the number of connections. Finally, we present a federated SRB service over MPI that allows fast TCP connections to stripe data across multiple server disks reaching 97% of the combined write capacity of multiple nodes.
Keywords :
client-server systems; computer networks; distributed shared memory systems; optimisation; pipeline processing; transport protocols; MPI; SDSC SRB; SDSC Storage Resource Broker; cluster-to-cluster throughput; commodity workstations; data grid infrastructure; disk I/O; fast TCP connections; federated SRB service; high-performance cluster storage server; multiple server disks; network-striped transport; pipeline optimizations; remote connection; Aggregates; Bandwidth; Geophysics computing; Grid computing; High performance computing; Internet; Middleware; Network servers; Throughput; Web server;
Conference_Titel :
High Performance Distributed Computing, 2002. HPDC-11 2002. Proceedings. 11th IEEE International Symposium on
Print_ISBN :
0-7695-1686-6
DOI :
10.1109/HPDC.2002.1029931
