IBM TotalStorage DS4000 Storage Considerations for SAS 9 on the IBM eServer p590
This storage whitepaper presents the results, and recommendations regarding various SAS 9.1 disk layouts tested at the Beaverton Test Labs. The storage disk layout was modified and monitored in concert with CPU sizing exercises by a combined SAS and IBM SAS International Competency Center (ICC) team. The test environment consisted of SAS 9.1 on the IBM® p590 eServer™ with AIX® 5.3 and the IBM TotalStorage DS4000. Workload for the testing was provided by the SAS Extract Transform Load (ETL) Test Suite. The SAS 9.1 hardware test environment and disk layouts tested during the benchmark are detailed in this document. Relative results of the storage testing are presented in this paper along with recommendations regarding the IBM TotalStorage DS4000 disk layout for SAS 9.1.
Website: www.sas.com | Filesize: 366kb
No of Page(s): 13
This document presents the storage findings and recommendations for SAS 9.1 with IBM TotalStorage DS4000 disk arrays (formerly known as FAStT). Recently IBM and SAS team members tested SAS 9.1 at the IBM Beaverton Test Labs. SAS ETL Test Suite 1.4 was used for the test workload generator. The test environment hardware included a SAN attached IBM TotalStorage DS4400 and the IBM p5 eServer using AIX. In addition to the DS4400, most of the storage findings and layout recommendations can be applied toward any of the members of the IBM TotalStorage DS4000 family including the DS4100, DS4300, DS4400, DS4500, and DS4800.
The objectives achieved in this test effort included CPU sizing exercises, and storage disk layout validation/performance testing. Storage testing focused on filesystem configuration, host bus adapters quantity, and disk array placement. In addition, the team tested the DS4000 Storage Management physical disk array placement software algorithm and completed the TotalStorage Proven™ certification testing for SAS 9 with the stated IBM hardware.
During testing the initial DS4000 storage I/O parameters were documented and used as a performance baseline. The team changed DS4000 disk parameters and tested those changes with the same ETL Test Suite data comparing the results against the baseline test. This document focuses on some of the relevant application performance results along with DS4000 storage configuration recommendations based on the completed performance testing.
The ETL Reference Architecture was exercised using the partitioning capabilities of a large 32-way IBM eServer p5-590. Four 2 Gbps Fiber channels adapters were attached to each partition and both ETL large and ETL Extra Large test scenarios were conducted against a 4-way, 8-way and 16-way partition connected via switched fiber channel to an IBM TotalStorage DS4400 Storage Server.
The DS4400 consisted of two controllers attached to 8 EXP700 expansion drawers fully populated with 112 36GB 15K RPM drives. The AIX RDAC (Redundant Disk Array Controller) software was installed in order to provide the disk array Fibre Channel path redundancy as well as disk array DS4400 controller redundancy.
Get the file Download here