Thanks Guy, that was the problem. Maybe I'm just a bit slow (today?), but I never saw anyplace in the FAQ or the howto's about the partition type needed. Just for mailing list postarity, (and for those trolling google looking for things like: kernel: md: personality 1 is not loaded! mdadm: RUN_ARRAY failed: Invalid argument Linux-RAID mkraid: aborted, see the syslog and /proc/mdstat for potential clues ) ALL RAID devices must have a partition type of FD BEFORE you run mkraid or mdadm. Now all that's left to do is format the bad boy, and off I go! Thanks agian! David ------- David J. Novak GSM Radio Firmware GSM Products Division CE/NSS Motorola Life v7.0 -------------------------------------------------------------------------- "Not all who wander are lost." - J.R.R. Tolkien -----Original Message----- From: Guy [mailto:bugzilla@watkins-home.com] Sent: Friday, April 02, 2004 2:24 PM To: 'Novak David-DNOVAK1' Subject: RE: Linear Raid Problems... The partitions should be type FD. No formatting required for the partitions. However, once you create the array, it must be formatted. The array will be /dev/md0 >From "man mdadm": -n, --raid-devices= number of active devices in array. My man page makes no reference to --raid-disks. However, I would have used "-n", so not sure which is correct. I guess mdadm or its man page needs some work! Try this: modprobe -v raid0 or this: modprobe -v linear Depending on which you use. Again, I have never needed to do this. Guy -----Original Message----- From: linux-raid-owner@vger.kernel.org [mailto:linux-raid-owner@vger.kernel.org] On Behalf Of Novak David-DNOVAK1 Sent: Friday, April 02, 2004 2:36 PM To: linux-raid@vger.kernel.org Subject: RE: Linear Raid Problems... Hmmm...Nope, mdadm didn't like --raid-devices. It looks like it's finding all the drives OK, but when it actually puts the device into service with the run command, that's when it's puking. Should I not have fdisk'ed the drives? I assumed that they needed SOME sort of formating. I just noticed this in my syslog: Apr 2 12:57:13 Homer kernel: md: bind<sdb1,1> Apr 2 12:57:13 Homer kernel: md: bind<sdc1,2> Apr 2 12:57:13 Homer kernel: md: bind<sdd1,3> Apr 2 12:57:13 Homer kernel: md: sdd1's event counter: 00000000 Apr 2 12:57:13 Homer kernel: md: sdc1's event counter: 00000000 Apr 2 12:57:13 Homer kernel: md: sdb1's event counter: 00000000 Apr 2 12:57:13 Homer kernel: md: personality 1 is not loaded! Apr 2 12:57:13 Homer kernel: md: md0 stopped. Apr 2 12:57:13 Homer kernel: md: unbind<sdd1,2> Apr 2 12:57:13 Homer kernel: md: export_rdev(sdd1) Apr 2 12:57:13 Homer kernel: md: unbind<sdc1,1> Apr 2 12:57:13 Homer kernel: md: export_rdev(sdc1) Apr 2 12:57:13 Homer kernel: md: unbind<sdb1,0> Apr 2 12:57:13 Homer kernel: md: export_rdev(sdb1) OK. So I'm 99.99% sure that the personality (linear) is not loaded. Question: Where is this personality located? Where do I get this personality from? How do I load this personality? Thanks again! David ------- David J. Novak GSM Radio Firmware GSM Products Division CE/NSS Motorola Life v7.0 -------------------------------------------------------------------------- "Not all who wander are lost." - J.R.R. Tolkien -----Original Message----- From: Guy [mailto:bugzilla@watkins-home.com] Sent: Friday, April 02, 2004 1:14 PM To: Novak David-DNOVAK1; linux-raid@vger.kernel.org Subject: RE: Linear Raid Problems... Personality... I have never needed to talk any steps to install personalities. Did you check the logs for any info? Maybe replace: --raid-disks=3 With --raid-devices=3 If this is the issue, maybe mdadm sucks! :) At least the error message sucks! RAID0 should give better performance over linear. With RAID0 the array will be striped across all 3 disks, as much as can be. With linear the 3 disks are used 1 after the other. Guy -----Original Message----- From: linux-raid-owner@vger.kernel.org [mailto:linux-raid-owner@vger.kernel.org] On Behalf Of Novak David-DNOVAK1 Sent: Friday, April 02, 2004 12:35 PM To: linux-raid@vger.kernel.org Subject: RE: Linear Raid Problems... Thanks for the quick response Guy, Here's some more info for you. Here's the output of the dd stuff: Homer:/# dd if=/dev/sdc of=/dev/null bs=64k 65533+1 records in 65533+1 records out Homer:/# dd if=/dev/sdb of=/dev/null bs=64k 138197+1 records in 138197+1 records out Homer:/# dd if=/dev/sdd of=/dev/null bs=64k 32773+1 records in 32773+1 records out That looks OK to me. So I added the mdadm package (so nobody would yell), and tried this: Homer:/# mdadm --create /dev/md0 --chunk=64 --level=linear --raid-disks=3 /dev/sdb1 /dev/sdc1 /dev/sdd1 mdadm: /dev/sdb1 appears to contain an ext2fs file system size=8841040K mtime=Thu Apr 1 18:22:11 2004 mdadm: /dev/sdc1 appears to contain an ext2fs file system size=4194156K mtime=Thu Apr 1 18:23:00 2004 mdadm: /dev/sdd1 appears to contain an ext2fs file system size=2076320K mtime=Thu Apr 1 18:23:19 2004 Continue creating array? y mdadm: RUN_ARRAY failed: Invalid argument Homer:/# Then: Homer:/# cat /proc/mdstat Personalities : read_ahead not set unused devices: <none> Homer:/# Do I not have the correct "personality" loaded? How do I check that? If I don't have it loaded, how do I load it? Thanks again! David ------- David J. Novak GSM Radio Firmware GSM Products Division CE/NSS Motorola Life v7.0 -------------------------------------------------------------------------- "Not all who wander are lost." - J.R.R. Tolkien -----Original Message----- From: Guy [mailto:bugzilla@watkins-home.com] Sent: Friday, April 02, 2004 8:55 AM To: Novak David-DNOVAK1; linux-raid@vger.kernel.org Subject: RE: Linear Raid Problems... Looks like you have a bad disk! Apr 2 07:47:39 Homer kernel: I/O error: dev 08:21, sector 23758976 I test my disks with the dd command. Examples: To test the whole disk: dd if=/dev/sdb of=/dev/null bs=64k To test one partition: dd if=/dev/sdb1 of=/dev/null bs=64k I would test the whole disk. It looks like your array is assembled, but not started. Try: raidstop /dev/md0 mkraid /dev/md0 You should consider switching from mkraid to mdadm. Most people here will YELL at you for using mkraid! :) For details: man mdadm In a related issue, you should consider using RAID5. You will lose the space of one disk. However, if a disk fails no data loss and the data is accessible. Oops... I see your disks have very different sizes. Don't use RAID5! Guy -----Original Message----- From: linux-raid-owner@vger.kernel.org [mailto:linux-raid-owner@vger.kernel.org] On Behalf Of Novak David-DNOVAK1 Sent: Friday, April 02, 2004 9:27 AM To: 'linux-raid@vger.kernel.org' Subject: Linear Raid Problems... Hi all, Sorry, but this message might be a tad long. But I'm running an UltraSparc Debian Linux (with SMP!). Anyway, I've got a few extra disks I want to span into one Uber-partion, and figured Software RAID (linear) was the way to go. I've read the howto's and the faq's, and thought I was doing everything right...but alas something is amiss, and I implore for the Linux community's help. Here's my setup: /etc/raidtab: raiddev /dev/md0 raid-level linear nr-raid-disks 3 chunk-size 32 persistent-superblock 1 device /dev/sdb1 raid-disk 0 device /dev/sdc1 raid-disk 1 device /dev/sdd1 raid-disk 2 mkraid command: Homer:/etc# mkraid /dev/md0 handling MD device /dev/md0 analyzing super-block disk 0: /dev/sdb1, 8841042kB, raid superblock at 8840960kB disk 1: /dev/sdc1, 11879560kB, raid superblock at 11879488kB mkraid: aborted, see the syslog and /proc/mdstat for potential clues. mkraid version: Homer:/etc# mkraid --version mkraid version 0.90.0 Kernel version: Homer:/etc# cat /proc/version Linux version 2.4.19 (root@hopper) (gcc version egcs-2.92.11 19980921 (gcc2 ss-9 80609 experimental)) #1 SMP Fri Oct 4 19:11:01 EDT 2002 After running mkraid they syslog says: Apr 2 07:47:39 Homer kernel: I/O error: dev 08:21, sector 23758976 In desperation tried: mkraid --really-force /dev/md0 DESTROYING the contents of /dev/md0 in 5 seconds, Ctrl-C if unsure! handling MD device /dev/md0 analyzing super-block disk 0: /dev/sdb1, 8841042kB, raid superblock at 8840960kB disk 1: /dev/sdc1, 11879560kB, raid superblock at 11879488kB disk 2: /dev/sdd1, 2076320kB, raid superblock at 2076224kB mkraid: aborted, see the syslog and /proc/mdstat for potential clues. Then syslog says: Apr 2 07:50:18 Homer kernel: md: array md0 already exists! Which is weird because that exact command didn't work last night... but now mdstat says: Homer:/etc# cat /proc/mdstat Personalities : read_ahead not set9L, 906C 1,1 All md0 : inactive sdd1[2] sdc1[1] sdb1[0] 0 blocks unused devices: <none> Which is weird, because last night it didn't say that... So I think the kernel thinks it's there right, but when I do an fdisk: Homer:/# fdisk /dev/md0 Unable to read /dev/md0 Or if I try to create a fs: Homer:/# mkfs.ext3 /dev/md0 mke2fs 1.27 (8-Mar-2002) mkfs.ext3: Device size reported to be zero. Invalid partition specified, or partition table wasn't reread after running fdisk, due to a modified partition being busy and in use. You may need to reboot to re-read your partition table. I think that's all the data somebody really smart might need to diagnose my problem. Thanks in advance for any info! David ------- David J. Novak GSM Radio Firmware GSM Products Division CE/NSS Motorola Life v7.0 -------------------------------------------------------------------------- "Not all who wander are lost." - J.R.R. Tolkien - 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 - 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
