Review the following considerations when using ZFS with a database product.
If the database uses a fixed disk block or record size for I/O, set the ZFS recordsize property to match it. You can do this on a per-file system basis, even though multiple file systems might share a single pool.
With ZFS's copy-on-write design, tuning down the recordsize is a way to improve OLTP performance at the expense of batch reporting queries.
ZFS checksums every block stored on disk. This alleviates the need for the database layer to checksum data an additional time. If checksums are computed by ZFS instead of at the database layer, any discrepancy can be caught and fixed before the data is returned to the application.
UFS direct I/O is used to overcome some of the design deficiencies of UFS and to eliminate double buffering of data. In ZFS, the UFS design deficiencies do not exist and ZFS uses the primarycache and secondarycache properties to manage buffering data in the ARC. Note that using the secondarycache (L2ARC) property to improve random reads also requires the primarycache property to be enabled.
Keep pool space under 90% utilization to maintain pool performance.