I set up a slog and cache device using slices on a 30g OCZ Vertex 2 about a year ago. In late August, I decided that it wasn't worth it to have the slog on a drive that might fail to honor cache flushes, especially since most of the writes I'm doing are via smb which don't need a slog as direly as nfs or iscsi.
Today I took a look at fmdump -e and saw this: Oct 04 02:00:49.3620 ereport.io.scsi.cmd.disk.tran Oct 04 02:05:59.5121 ereport.fs.zfs.probe_failure fmdump shows: TIME UUID SUNW-MSG-ID Oct 04 02:06:15.4763 0247d84c-22a1-660a-a065-bfecf588abc8 ZFS-8000-FD and zpool status shows the cache device as FAULTED. My takeaway? A small MLC device will get eaten alive when used as a slog device. Even used as a cache, it'll wear out pretty quickly. -B -- Brandon High : bh...@freaks.com _______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
