[Bug 777067] New: none of 44 existing partitions on sdc displayed/available
https://bugzilla.novell.com/show_bug.cgi?id=777067 https://bugzilla.novell.com/show_bug.cgi?id=777067#c0 Summary: none of 44 existing partitions on sdc displayed/available Classification: openSUSE Product: openSUSE 12.2 Version: RC 2 Platform: x86-64 OS/Version: Other Status: NEW Severity: Major Priority: P5 - None Component: Installation AssignedTo: bnc-team-screening@forge.provo.novell.com ReportedBy: mrmazda@earthlink.net QAContact: jsrain@suse.com Found By: --- Blocker: --- Created an attachment (id=503198) --> (http://bugzilla.novell.com/attachment.cgi?id=503198) y2logs I tried to HTTP install from distribution/12.2 loading initrd & kernel with Grub from ext2 partition sda1 routinely used for booting 3 openSUSE installations on multiple md RAID1 partitions created across sda & sdb. After selecting custom partitioning for experts, the partitioner shows 500GB sdc as having no partitions, even though the last on it is sdc44, confirmed via fdisk -l /dev/sdc on tty2. No partitions on sdc are any part of any RAID. sdc has mostly ever been used as a backup device. Since the installation target for the exercise is the existing sdc15, installation is apparently impossible. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=777067 https://bugzilla.novell.com/show_bug.cgi?id=777067#c1 --- Comment #1 from Felix Miata <mrmazda@earthlink.net> 2012-08-23 18:18:03 UTC --- After reducing partition count to 35, which alone did not help on next try, I tried again with BrokenModules=hptiop,hpt37x,scsi_dh,raid0,raid10,raid456,raid6_pq,dm_mod,dm_multipat and got what appeared to be a complete minimal installation + apache + samba/cifs + nfs-kernel-server. Y2logs for the subsequent install available on request. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=777067 https://bugzilla.novell.com/show_bug.cgi?id=777067#c2 Christian Boltz <suse-beta@cboltz.de> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |suse-beta@cboltz.de AssignedTo|bnc-team-screening@forge.pr |aschnell@suse.com |ovo.novell.com | --- Comment #2 from Christian Boltz <suse-beta@cboltz.de> 2012-08-23 21:33:30 CEST --- Assigning to Arvin (maintainer of yast2-storage). Arvin, if you think another part of YaST is responsible for this bug, please reassign ;-) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=777067 https://bugzilla.novell.com/show_bug.cgi?id=777067#c3 Arvin Schnell <aschnell@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |aschnell@suse.com AssignedTo|aschnell@suse.com |nfbrown@suse.com --- Comment #3 from Arvin Schnell <aschnell@suse.com> 2012-08-24 08:12:38 UTC --- The disk sdc is part of a BIOS RAID: /sbin/dmraid -s -c -c -c hpt37x_ccedfbhiib:156301440:128:mirror:ok:0:1:0 /dev/sdc:hpt37x:hpt37x_ccedfbhiib:mirror:ok:156301477:10 But the RAID is broken: /sbin/dmraid -ay -p ERROR: creating degraded mirror mapping for "hpt37x_ccedfbhiib" RAID set "hpt37x_ccedfbhiib" was activated RAID set "hpt37x_ccedfbhiib" was not activated So for YaST sdc is used by the RAID and its partitions are not shown. Why the RAID is broken or why dmraid doesn't setup device-mapper tables for the partitions I cannot say. Adding hpt37x to broken modules disables the BIOS RAID completely and the system behaves like you expected. Anyway, removing the BIOS RAID signature from the disk is a better solution. Neil, can you comment on the behavior of dmraid? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=777067 https://bugzilla.novell.com/show_bug.cgi?id=777067#c4 Neil Brown <nfbrown@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO InfoProvider| |mrmazda@earthlink.net --- Comment #4 from Neil Brown <nfbrown@suse.com> 2012-08-24 09:15:17 UTC ---
Neil, can you comment on the behavior of dmraid?
I'd have to dig into the code and see if I found something... Felix, if you haven't already removed the BIOS RAID signatures, could you run dmraid -r -D /dev/sdc which should create a directory called "dmraid.hpt37x" or something like that. Then tar up that directory and attach it? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=777067 https://bugzilla.novell.com/show_bug.cgi?id=777067#c5 Felix Miata <mrmazda@earthlink.net> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW InfoProvider|mrmazda@earthlink.net | --- Comment #5 from Felix Miata <mrmazda@earthlink.net> 2012-08-24 10:58:22 UTC --- Created an attachment (id=503405) --> (http://bugzilla.novell.com/attachment.cgi?id=503405) tar -cvf of dmraid -r -D /dev/sdc output As I wrote in comment 0, this sdc has mostly ever been used as a backup device, and likely was never booted except to its DOS partition until attempting to put 12.2 on it. IIRC, this SATA device was initialized by cloning from an 80G PATA disk that had once been installed in a system that did have a hpt3xx controller enabled for fake RAID, but there have been no working systems with hpt3xx controllers here in many many moons. I've now wiped 0x3d sectors starting @ 0x01 on sdc. Could YaST have done this for me while I was attempting to install? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=777067 https://bugzilla.novell.com/show_bug.cgi?id=777067#c6 Neil Brown <nfbrown@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |nfbrown@suse.com AssignedTo|nfbrown@suse.com |aschnell@suse.com --- Comment #6 from Neil Brown <nfbrown@suse.com> 2012-09-17 01:52:20 UTC --- The "was not activated" looks like a fairly uninteresting bug (which I'll have a little look for) - the array actually is activated. I experimented a bit and I it quite easy - in the export/custom mode - to delete the HPT raid device, and then all the partitions appear and can be used. So I think that the real problem is that it isn't visually obvious that the "DM RAID" device is using the "sdc" device, and that by deleting it, other uses can be made of the "sdc" device. If you go to the "device graph" this becomes fairly obvious, But would you think of going there? Maybe if the "mountpoint" column were called "Used By", and the the 'sdc' device had "DM RAID" listed there, it might be more obvious.?? Up to you, Arvin, whether you think this can be improved at all. But 'dmraid' seem to be doing close enough to the correct thing that I don't think it is at fault here. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=777067 https://bugzilla.novell.com/show_bug.cgi?id=777067#c7 Arvin Schnell <aschnell@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution| |INVALID --- Comment #7 from Arvin Schnell <aschnell@suse.com> 2012-10-02 11:56:07 UTC --- Ok, what might have confused me is that although it's a mirror RAID parted detects no partitions. But that's due to an offset on the RAID: /sbin/dmsetup table 'hpt37x_ccedfbhiib' 0 156301477 linear 8:32 10 Assuming this is correct there is indeed nothing wrong. The user is also able to delete the RAID in YaST as Neil mentions. YaST has a column names "Used By" and the overview for sdc also shows the "used by" information. All operations on sdc also also display a popup that the disk is in use. So I don't see room for improvement. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply＠novell.com