On Thu, Feb 11, 2010 at 2:46 PM, Dan Pritts <da...@internet2.edu> wrote: > I've been considering it, but I talked to a colleague at another > institution who had some really awful tales to tell about promise > FC arrays. They were clearly not ready for prime time. > > OTOH a SAS jbod is a lot less complicated.
We have used two Promise Vtrak m500f fibre channel arrays in heavy I/O applications for several years. They don't always handle failing disks gracefully-- sometimes requiring a hard reboot to recover. This is partly due to crappy disks with weird failure modes. But a RAID system should never require a reboot to recover from a single disk failure. That defeats the whole purpose of RAID, which is supposed to survive disk failures through redundancy. However, Promise iSCSI and JBOD systems (we own one of each) are more stable. We use them with Linux (XFS) and OpenSolaris (ZFS), and haven't any experienced problems to date. Their JBOD systems, when filled with Western Digital RE3 disks, are an extremely reliable, low-cost, high performance ZFS storage solution. Daniel Bakken Systems Administrator Economic Modeling Specialists Inc 1187 Alturas Drive Moscow, Idaho 83843 _______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
