On Mon, Apr 24, 2006 at 07:45:27AM +1000, Neil Brown wrote: > your array isn't degraded. In this case it is (I think) very unusual > and may not be the cause of your corruption, but you should avoid > using the flag anyway. thanks a lot for your time and your attention, Neil. Your support it's fast and valuable, as usual. well, I wasted lot of hours, after my post, trying to find the reason of the corruptions I've got. Well, the problem is funny... I mean... I can cp hundred of giga, in ext2, without complain in dmesg/log, but if I umount the fs and run fsck I've got a lot of incredible problem (duplicated blocks, and so on). with ext3 it can works for hours, seldom I've got ext3-journal corruption. anyway, after fsck, the checksum of files is always good, and lost+found full of monster (some files need debugs to be eliminated (lsattr/chattr failed working with them)). after checking hardware, changing controllers, now I have changed even hd cables. at home I will re-run all the tests. I don't think it's a problem of raid software, of course. > > > > b) dm-encrypt /dev/md1 > > > > c) create fs with: > > mkfs.ext3 -O dir_index -L 'tritone' -i 256000 /dev/mapper/raidone > > > > d) export it via nfs (mounting /dev/mapper/raidone as ext2) > ^^^^ > > Why not ext3? Well, because I had to clone 1,5 TB of data, spread over a lot of disks, in one shot, and to avoid journal seeks I've done so. > > > > > e) start to cp-ing files > > > > f) after 1 TB of written data, with no problem/warning, one of the > > not-in-raid-array HD freeze > > This could signal a bad controller. If it does, then you cannot trust > any drives. well, it was my fault... I mean, I've got a Dell server, without enough internal room for all the disks. The source disk was out of the server, and I move it... it wasn't happy... anyway, I'm using HPT ATA PCI controller (well tested, I mean, I used the ones in the server since 2000). btw, 5 disks Maxtor, 500Giga each one. The problem isn't MD related, but it's the first time I've got so much problems finding the culprit of data corruption. Usually it's RAM/CPU fault, few times I've got problem with controller... but this time I'm going slightly mad... also, why meta and not data (file are checked with a stupid python script I wrote)... is there an ATA command triggered only with metadata? uhm... maybe mounting the array in synchronous mode I could gather more info, uhm... at the end, Neil, thanks a lot for your work. If you'll be in Italy, some day, I'll be happy to be your host. ciao, gelma - To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
