For anyone that is interested, here's a progress report. I created a new pool with only one mirror vdev of 2 disks, namely with the new SAMSUNG HD204UI. These drives, along with the older HD203WI, use Advanced Format Technology (e.g. 4K sectors). Only these drives had hard errors in my pool, as opposed the the old Seagates and WDs.
To create the new pool, I recompiled the zpool cmd to give the value of ashift 12 so that the new pool has an alignement of 4K instead of 512 bytes (see here : http://www.solarismen.de/archives/5-Solaris-and-the-new-4K-Sector-Disks-e.g.-WDxxEARS-Part-2.html). So I filled this new 4K aligned pool with 1.5TB of data, scrubbed it and no errors. I checked the log and no hard errors either. Usually after a scrub I get some hard errors. Maybe the pool needs to have more vdevs in it to really stress the HBA and produce hard errors, but it's a strange coincidence nonetheless that only the 4K drives had errors and then when used in a 4K aligned pool, no more errors. I'll probably re-create my original pool with only 4K drives in a 4K aligned pool and see what happens. -- This message posted from opensolaris.org _______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
