As recently reminded myself by accidently mixing up motherboards and getting a silicon image based SATA board those companies chipsets are to be avoided at all costs. There are currently 3 chipsets that I would recommend. Promise SATA (non RAID like the SATA TX4 and SATAII TX4) AHCI based chipsets (like Intels later chipsets which have published interfaces) 3ware SATA boards (8000, 9000 series) which have had drivers for a long time, these are "hardware raid" cards I have used promose and 3ware, and according to the libata docs the Intel stuff is good, if it is the new stuff. Older stuff is not so good. Terrence C.M. Connelly wrote: >-----BEGIN PGP SIGNED MESSAGE----- >Hash: SHA1 > > >I'm running an array with an Adaptec 2410SA card (two-port SATA). >It seems to run fine under CentOS, but I'm not booting from it (I >have a separate SCSI drive that has the OS installed on it; the >array is used for backups). I have lost the contents of the array >once when one of the drives failed (that was under Red Hat Linux >9, with the Adaptec driver package). I haven't been able to get >the current Adaptec driver package to install under CentOS, but as >the card seems to work fine without it, I haven't tried very hard. > > >I had some issues with the SATA cables included with the Adaptec >card and ended up replacing them with cables that seems to seat >better and wobble less. > > >One card I would definitely *not* recommend is Adaptec's >AAR-1210SA card, a two-port SATA card with a Silicon Image >chipset. After having lots of problems under RHL 9, I managed to >get it working with a Debian system with a custom 2.6 kernel. >Definitely not worth the money. > >Both of these arrays were meant as quickie solutions to relatively >unimportant problems. If I were to do them again, or do a more >serious project, I would definitely go with the 3Ware cards over >the Adaptec cards. I also had a very bad experience with a >Promise external RAID enclosure that has put me off Promise for >life. > >(Actually, I would seriously consider adding disk space via one or >more of Apple's Xserve RAID subsystems, as well.) > >YMMV. > > Claire > >*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* > Claire Connelly cmc at math.hmc.edu > Systems Administrator (909) 621-8754 > Department of Mathematics Harvey Mudd College >*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-*-* >-----BEGIN PGP SIGNATURE----- >Version: GnuPG v1.2.1 (GNU/Linux) >Comment: Processed by Mailcrypt 3.5.8 <http://mailcrypt.sourceforge.net/> > >iD8DBQFBlRZpB0pE8d7vd8wRAhAtAKCIsuxhinGoRTMqXNdKDXrthX/LiwCeMOwu >OyTaL0+Ti6OVMC6WTAVl2bk= >=GJZ2 >-----END PGP SIGNATURE----- >_______________________________________________ >CentOS mailing list >CentOS at caosity.org >http://lists.caosity.org/mailman/listinfo/centos > > >