Oracle ZFS Storage ZS3-4 Produces Best 2-Node Performance on SPECsfs2008 NFSv3
By Brian-Oracle on Sep 10, 2013
The Oracle ZFS Storage ZS3-4 storage system delivered world record two-node performance on the SPECsfs2008 NFSv3 benchmark, beating results published on NetApp's dual-controller and four-node high-end FAS6240 storage systems.
The Oracle ZFS Storage ZS3-4 storage system delivered a world record two-node result of 450,702 SPECsfs2008_nfs.v3 Ops/sec with an Overall Response Time (ORT) of 0.70 msec on the SPECsfs2008 NFSv3 benchmark.
The Oracle ZFS Storage ZS3-4 storage system delivered 2.4x higher throughput than the dual-controller NetApp FAS6240 and 4.5x higher throughput than the dual-controller NetApp FAS3270 on the SPECsfs2008_nfs.v3 benchmark at less than half the list price of either result.
The Oracle ZFS Storage ZS3-4 storage system had 42 percent higher throughput than the four-node NetApp FAS6240 on the SPECsfs2008 NFSv3 benchmark.
The Oracle ZFS Storage ZS3-4 storage aystem has 54 percent better Overall Response Time than the 4-node NetApp FAS6240 on the SPECsfs2008 NFSv3 benchmark.
Two node results for SPECsfs2008_nfs.v3 presented (in decreasing SPECsfs2008_nfs.v3 Ops/sec order) along with other select results.
Throughput SPECsfs2008_nfs.v3 Ops/sec — the Performance Metric
Overall Response Time — the corresponding Response Time Metric
Nodes — Nodes and Controllers are being used interchangeably
Complete SPECsfs2008 benchmark results may be found at http://www.spec.org/sfs2008/results/sfs2008.html.
2 TB memory
2 x 10GbE NICs
2 x shelves with 20 x 300 GB 15K RPM SAS-2 drives and 8 x 73 GB SAS-2 flash-enabled write-cache
SPECsfs2008 is the latest version of the Standard Performance Evaluation Corporation (SPEC) benchmark suite measuring file server throughput and response time, providing a standardized method for comparing performance across different vendor platforms. SPECsfs2008 results summarize the server's capabilities with respect to the number of operations that can be handled per second, as well as the overall latency of the operations. The suite is a follow-on to the SFS97_R1 benchmark, adding a CIFS workload, an updated NFSv3 workload, support for additional client platforms, and a new test harness and reporting/submission framework.
- Standard Performance Evaluation Corporation (SPEC) Home Page
- Oracle ZFS Storage ZS3-4
SPEC and SPECsfs are registered trademarks of Standard Performance Evaluation Corporation (SPEC). Results as of September 10, 2013, for more information see www.spec.org. Oracle ZFS Storage ZS3-4 Appliance 450,702 SPECsfs2008_nfs.v3 Ops/sec, 0.70 msec ORT, NetApp Data ONTAP 8.1 Cluster-Mode (4-node FAS6240) 260,388 SPECsfs2008_nfs.v3 Ops/Sec, 1.53 msec ORT, NetApp FAS6240 190,675 SPECsfs2008_nfs.v3 Ops/Sec, 1.17 msec ORT. NetApp FAS3270 101,183 SPECsfs2008_nfs.v3 Ops/Sec, 1.66 msec ORT.
Nodes refer to the item in the SPECsfs2008 disclosed Configuration Bill of Materials that have the Processing Elements that perform the NFS Processing Function. These are the first item listed in each of disclosed Configuration Bill of Materials except for EMC where it is both the first and third items listed, and HP, where it is the second item listed as Blade Servers. The number of nodes is from the QTY disclosed in the Configuration Bill of Materials as described above. Configuration Bill of Materials list price for Oracle result of US$ 423,644. Configuration Bill of Materials list price for NetApp FAS3270 result of US$ 1,215,290. Configuration Bill of Materials list price for NetApp FAS6240 result of US$ 1,028,118. Oracle pricing from https://shop.oracle.com/pls/ostore/f?p=dstore:home:0, traverse to "Storage and Tape" and then to "NAS Storage". NetApp's pricing from http://www.netapp.com/us/media/na-list-usd-netapp-custom-state-new-discounts.html.