Hi Mikael, Glad you got some feedback, yes the 80 gig Maxtors are unreliable. Out of fourty disks I've had 12 RMAs so far. It does seem that some have different usable capacities. If I force the geometry of a "smaller" one to the geometry of a more normal looking variety I start getting unrecoverable errors during a RAID-1 or 5 resync towards the end of the partition. I'm curious, though I don't have any data, if one of the variants is less reliable than the others. At any rate it's good that you're using a mirror, happy RAIDing! -Kanoa Mikael Johansson wrote: > >Hello All! > >Thanks for all the answers, they helped a lot. Some comments below: > > >On Sat, 9 Feb 2002, Danilo Godec wrote: >======================================= > >>You can use the 'hdX=cylinders,heads,sectors' kernel parameter to make the >>kernel see the second drive the way you want it to see (I use this a >>lot) ie. identical to the first one. >> > >Hmm, it does sound a bit unsafe, one boot (from a CD, say) with the wrong >parameters and everything's screwed, right? > >>I don't belleive any performance loss would result in using different >>partitions. >> > >Good, I guess I'll go with that then! > >>And using RAID0 for swap is not quite reasonable - you don't gain any >>performance boost as swap code already does IO balancing between two (or >>more) swap partitions, if you set them to the same priority. >> > >OK, I've tried to figure this out before also, but I haven't found a >concensus on which scheme is better. But I guess the kernel in principle >should be able to perform the balancing better than using a fixed chunk >size. Thanks. > > >On Sat, 9 Feb 2002, Alvin Oga wrote: >==================================== > >>lots of identical drives gets different number of chs for >>hda vs hdb on different motherboards but is usally identical between hda >>and hdc and same between hdb and hdd >> >>so if you have a 2-drive raid1 .. just use hda and hdc as the >>preferred configuration .. >> > >My disks are located on hda and hdc, so the above seems not to apply in my >case. There is a difference on the buses though, I have a CD-drive on hdd, >while hdb is empty. Maybe that could explain why the identification goes >bad? > > >On Sat, 9 Feb 2002, Kanoa Withington wrote: >=========================================== > >>Some of the Maxtor 80gig disks actually are different. I have nearly >>forty of them and I've found at least three different varieties. My >>experiences with this particular model of disk are generally >>frustrating, though other Maxtor models have been very consistent and >>reliable. Aside from your quest to force the geometry and make it stick, >>which is necessary in your situation, you should not assume that the >>drives actually are identical just because they have the same sticker on >>them. The only thing that has worked for me is to partition them to the >>lowest common demoninator, which is using only 78 gigabytes from the >>beginning of the disk. I don't know if that is the highest common >>capacity but it's one that works for me and it might help you too. >> > >That's interesting. Do the varieties have slight differences in capacity? >Because these two in the end have the exact same amount of bytes >available. > >Does the "frustration" arise from only this inconsistency, or are there >other things wrong with the combination Linux/Maxtor 80GB? > > >Thanks to all again, have a nice day, > Mikael J. > >- >To unsubscribe from this list: send the line "unsubscribe linux-raid" in >the body of a message to majordomo@vger.kernel.org >More majordomo info at http://vger.kernel.org/majordomo-info.html > > - To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
