ZFS on RBODs - Leveraging RAID Controllers for Metrics and Enclosure Management

Stearman, D. M.¹ 

[1]Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)

关键词: MANAGEMENT;    METRICS;    MONITORING;    PERFORMANCE;    COMPUTER CODES;    CORRECTIONS;    COST;    ERRORS;    EQUIPMENT INTERFACES;    MONITORING;    MAGNETIC DISKS;   

DOI  :  10.2172/1184101
RP-ID  :  LLNL-TR--669183
PID  :  OSTI ID: 1184101
Others  :  TRN: US1500303

学科分类：数学（综合）

美国|英语

来源: SciTech Connect

【 摘 要 】

Traditionally, the Lustre file system has relied on the ldiskfs file system with reliable RAID (Redundant Array of Independent Disks) storage underneath. As of Lustre 2.4, ZFS was added as a backend file system, with built-in software RAID, thereby removing the need of expensive RAID controllers. ZFS was designed to work with JBOD (Just a Bunch Of Disks) storage enclosures under the Solaris Operating System, which provided a rich device management system. Long time users of the Lustre file system have relied on the RAID controllers to provide metrics and enclosure monitoring and management services, with rich APIs and command line interfaces. This paper will study a hybrid approach using an advanced full featured RAID enclosure which is presented to the host as a JBOD, This RBOD (RAIDed Bunch Of Disks) allows ZFS to do the RAID protection and error correction, while the RAID controller handles management of the disks and monitors the enclosure. It was hoped that the value of the RAID controller features would offset the additional cost, and that performance would not suffer in this mode. The test results revealed that the hybrid RBOD approach did suffer reduced performance.