By Parnian Taidi-Oracle on May 06, 2016
SAP Adaptive Server Enterprise (ASE) database 16.0 and 15.7 are certified to run on Oracle Solaris 11 systems using ZFS as data storage and logs. Recent testing with OLTP workloads on Oracle SPARC systems with Solaris 11 show better or on-par performance with ZFS compared to UFS or raw devices.
Oracle Solaris ZFS is a powerful file system that combines file system capabilities with storage features that are traditionally offered by a volume manager. ZFS is the default file system in Oracle Solaris 11 and includes integrated data services, such as compression, encryption, and snapshot and cloning. Oracle Solaris and ZFS are also the foundation of Oracle ZFS Storage Appliance.
ZFS Best Practices For SAP ASE Database:
- Create separate zpools for data and the transaction log.
- Create zpool for data using multiple LUNs for better I/O bandwidth.
- ZFS record size for data is one of key parameters for optimal performance. The default value of 128K is too large and causes page overhead. For Solaris SPARC systems with the DB page size of 8K or less, a ZFS record size of 32K provides the best performance for both file system block and raw zvol.
- Place ZIL (ZFS Intent Log) on SSDs for improved latency of both data and transaction log.