Note: This is a public test instance of Red Hat Bugzilla. The data contained within is a snapshot of the live data so any changes you make will not be reflected in the production Bugzilla. Email is disabled so feel free to test any aspect of the site that you want. File any problems you find or give feedback at bugzilla.redhat.com.
Bug 113201
Summary: | Partition table is inconsistent warnings | ||
---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | Mike McLean <mikem> |
Component: | parted | Assignee: | Jeremy Katz <katzj> |
Status: | CLOSED DUPLICATE | QA Contact: | |
Severity: | medium | Docs Contact: | |
Priority: | medium | ||
Version: | rawhide | CC: | ccradu, chabotc, d.bz-redhat, gczarcinski, jaaksimm, katzj, leonard-rh-bugzilla, marius.andreiana, mattdm, p.van.egdom, rdieter, t.matsuu |
Target Milestone: | --- | ||
Target Release: | --- | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Fixed In Version: | Doc Type: | Bug Fix | |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2006-02-21 19:00:41 UTC | Type: | --- |
Regression: | --- | Mount Type: | --- |
Documentation: | --- | CRM: | |
Verified Versions: | Category: | --- | |
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
Cloudforms Team: | --- | Target Upstream Version: | |
Embargoed: |
Description
Mike McLean
2004-01-09 17:32:13 UTC
This is still happening lots. It interferes with automated testing so any help would be appreciated. :) Still happening regularly on a couple machines. Hmmm, not happening with any of the boxes I have. Can you get a box that hits the consistently installed and let me log into it? I am also seeing this problem regularly on one system. Unfortunately, it would be impractical to configure things so that you could login. However, I am willing to collect any information (try anything) that would help. What is really strange is that the system has two Maxtor 60GB drives (ST060H6) as hda and hdb with a Seagate 20GB drive (ST320420A) as hdc. The hda and hdc drives use the wrong (non lba) parameters but the hdb drive is fine. This is during the FC2-T1 install. Post install, running parted on hdb or hdc do not produce errors but on hda I get "... incorrect partition table ... Ignore/Cancel?". Running fdisk has no problem and shows lba values. -------------------------------------------------------------------- Here is the output of parted for hda: parted /dev/hda print Warning: Unable to align partition properly. This probably means that another partitioning tool generated an incorrect partition table, because it didn't havethe correct BIOS geometry. It is safe to ignore,but ignoring may cause (fixable) problems with some boot loaders. Ignore/Cancel? I Disk geometry for /dev/hda: 0.000-58644.140 megabytes Disk label type: msdos Minor Start End Type Filesystem Flags 1 0.031 10001.403 primary fat32 boot, lba 2 10001.404 11962.463 primary linux-swap 3 11962.463 12942.993 primary ext3 4 12942.993 58643.525 extended 5 12943.024 22489.431 logical ext3 6 22489.462 32035.869 logical ext3 7 32035.900 41582.307 logical ext3 8 41582.338 51128.745 logical ext3 9 51128.776 58643.525 logical ext2 Information: Don't forget to update /etc/fstab, if necessary. -------------------------------------------------------------------- and here is fdisk: fdisk -l /dev/hda Disk /dev/hda: 61.4 GB, 61492838400 bytes 255 heads, 63 sectors/track, 7476 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start End Blocks Id System /dev/hda1 * 1 1275 10241406 c W95 FAT32 (LBA) /dev/hda2 1276 1525 2008125 82 Linux swap /dev/hda3 1526 1650 1004062+ 83 Linux /dev/hda4 1651 7476 46797345 5 Extended /dev/hda5 1651 2867 9775521 83 Linux /dev/hda6 2868 4084 9775521 83 Linux /dev/hda7 4085 5301 9775521 83 Linux /dev/hda8 5302 6518 9775521 83 Linux /dev/hda9 6519 7476 7695103+ 83 Linux --------------------------------------------------------------------- Win2k was installed first into hda1. Then hda2 and hda3 were created as a small FC1 linux system (which saw everything OK as LBA). The remaining partitions were defined/allocated and initially formated under that FC1 system. OK, I have taken the two versions of parted (1.6.3-31 and 1.6.3-33) and built the packages on both a FC1 (2.4 kernel) and FC1-T1 (2.6 kernel). On FC1 (2.4 kernel) both versions of parted produce the same results (nor error messages). On FC2-T1 (2.6 kernel), they both produce amn error message about "Unable to align partition properly" for hda. For hdb, parted-1.6.3-31 produces the suggested geometry message but parted-1.6.3-33 does not since it was commented out. I am coming to the conclusion that this is a kernel problem and not parted (although parted may be sensitive to something that fdisk is not). I am seeing exactly the same thing, however, I cannot boot into my WinXP partition since installing Fedora Core 2 test 1!! I can still mount the partition from Linux, and the first sector looks valid (eg Jump command, "NTFS " etc). Same here, the error occurs when installing FC2-test1. The installer reports the warning "Unable to align partition properly. This probably means that another partitioning tool generated an incorrect partition table...". I have done 6-8 installation on my 80G Maxtor 6Y080 model, none successful. The machine does not boot after the installation (bios message that it is "missing operating system"). I have installed several ways: - deleted all partitions and started from scratch. - installed it in side with windows ntfs partition. After installation the partition table was in a bad state. I used Partition Magic to check the situation. In some cases it tried fixing some partition values, in other cases, recommended to re-create all partitions, due to bad drive geometry. This seems to be a bug in the FC2 tools. The Linux kernel no longer does partition guessing (its a heuristic and policy at best), as a result the parted tools should be honouring existing partition table claims when they are present. Failure to do so causes very bad things to happen. Previously these situations the kernel itself would report the partition table or BIOS guess it made, now its firmly in userspace. Is this bug fixed (or addressed) in FC2 test2 ? I do not believe it is fixed. This is a problem with parted and that has not changed from what was/is in the development tree as of a couple of days ago (which should be pretty much what is in Test2). See the "Unable to align partition" discussion on the fedora-test-list mailing list from yesterday (Mar 28 2004). This problem still occurs, but *much* less often than it did. When I first filed it, nearly all my testmachines would hit it. Now only a couple of those machines ever hit the bug. I have a system which seems to hit it every time (dual boot with Win2k installed in first partition on hda). I will be installing FC2 Test2 on that system later this week. We will see. Just de-partitioned my drive and the installation went ahead without a problem. At first tried removing just the linux partition and leaving the vfat. Only installed when all partitions removed. System: Shuttle SN41G2 - Athlon 2400+ Motherboard: Shuttle FN41 (bios version is FN41S025) Hard disk: Seagate 120 GB, model ST380021A with 8MB of cache Depending on the "Access Mode" used, my BIOS detect the following geometry for the HD: o "Access Mode" on "Auto" o "Access Mode" on "LBA" Cylinder = 38309 Cylinder = 9729 Head = 16 Head = 255 Sector = 255 Sector = 63 1. BEFORE installing FC2test2, I boot in FC1 and I get the following : # hdparm -g /dev/hda geometry = 9729/255/63 # fdisk /dev/hda" 255 heads, 63 sectors/track, 9729 cyl. NB : I get the same result whatever the "Access Mode" is ("Auto" or "LBA"). 2. DURING install of FC2T2, I get the following warning : "Unable to align partition properly. This probably means that another partioning tool generated an incorrect partition table, because it didn't have the correct BIOS geometry. It is safe to ignore, but ignoring may cause (fixable) problems with some boot loaders". After that DiskDruid displays a strange partition tables containing some small free spaces (< 1 cyl. each) beetween my partitions. NB : This happens with both "Access Mode" on "Auto" and "Access Mode" on "LBA". But this doesn't happen when installing FC1 on the same HD, with the same partition table ! 3. AFTER installing FC2T2, detected disk geometry is different : - If I boot in FC2T2 : # hdparm -g /dev/hda geometry = 65535/16/63 (with "Access Mode" on "Auto" or "LBA") # parted /dev/hda print Warning: Unable to align partition properly... - If I boot in FC1 : # hdparm -g /dev/hda geometry = 155061/16/63 (with "Access Mode" on "Auto") geometry = 9729/255/63 (with "Access Mode" on "LBA") # parted /dev/hda print No warning. - With both FC1 and FC2test2, "fdisk /dev/hda" gives : 16 heads, 63 sectors/track, 155061 cyl. And with "fdisk -H 255 /dev/hda", I get : 255 heads, 63 sectors/track, 9729 cyl. But this time, if I print the partition table with "p", I get "Partition 1 does not end on cylinder boundary" and the same message for partitions 2, 3 and 4. 4. With "Access Mode" set to "Auto", Windows XP (/dev/hda1) doesn't boot anymore. (It did boot before installing FC2T2). With "Access Mode" set to "LBA", it does boot again. Still broken in FC2T2 due to parted. fdisk gets it right. Parted needs to trust the on disk partition data. This is not a dual boot with Windows but I have on different partitions of the same system RH9 and FC2. Running 'parted' from RH9 results in the following whine: Error: The partition table on /dev/hdb is inconsistent. There are many reasons why this might be the case. However, the most likely reason is that Linux detected the BIOS geometry for /dev/hdb incorrectly. GNU Parted suspects the real geometry should be 39813/16/63 (not 2498/255/63)..... The one from FC2T2 is less informative: Warning: Unable to align partition properly. This probably means that another partitioning tool generated an incorrect partition table, because it didn't havethe correct BIOS geometry. It is safe to ignore,but ignoring may cause (fixable) problems with some boot loaders. Ignore/Cancel? Both versions of 'parted' print the same partition table for hda: Disk geometry for /dev/hda: 0.000-19595.948 megabytes Disk label type: msdos Minor Start End Type Filesystem Flags 1 0.031 1000.125 primary ext3 boot 2 1000.125 19595.460 extended 5 1000.156 7000.382 logical ext3 6 7000.414 9000.632 logical linux-swap 7 9000.664 16425.834 logical ext3 8 16425.866 19595.460 logical ext3 with 'fdisk' in both cases giving for the same disk: Disk /dev/hda: 20.5 GB, 20547841536 bytes 16 heads, 63 sectors/track, 39813 cylinders Units = cylinders of 1008 * 512 = 516096 bytes Device Boot Start End Blocks Id System /dev/hda1 * 1 2032 1024096+ 83 Linux /dev/hda2 2033 39813 19041624 5 Extended /dev/hda5 2033 14223 6144232+ 83 Linux /dev/hda6 14224 18287 2048224+ 82 Linux swap /dev/hda7 18288 33374 7603375+ 83 Linux /dev/hda8 33374 39813 3245665+ 83 Linux Note this "geometry" which, according to parted from RH9, "should be". During an installation anaconda, obviously, came up with some dire warnings about partitioning I created, and which I totally ignored, and everything was happy after that. Issue still reproduces (for me) with Fedora Core 2 Test 3. Existing Install of Windows 2000. Attempting to install Fedora onto second partition. Installer Warns: Unable to Align Partition......(I truncated the error message for this comment. The message appeared to be the same as the error message above) Since I had to FDISK my machine to recover from this bug in FC2T2, I am going to stop trying to test FC2T3. Is this an issue that people hope to have fixed for FC2 GOLD? Is there some work around for people that have an existing partition structure? Same here, I confirm the error (Unable to Align Partitions) still occurs for me as well. I tried installing FC2-test3 aside Windows XP ntfs partition. Also the bug still destroyes the partition data, since I lost a partition when trying to install test3. This bug should have high priority. Though the installer crashed on me. For more details see the following bug: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=121798 Same problem here! "Unable to align..." - also with FCore Test 3. As a test: I started installation with older FC1 Release and created all partitions without problems. Then restarted installation with FC2 - Test 3 and tried to hang in the previously created partitions - but same problem again. Any fix or workaround for now? I guess another data point won't hurt. I am also dual booting with win2k. Two disk drives Maxtor 40G and Maxtor 80g. Slightly different results with test2 and test3. In both cases I installed everything to an existing hda6, even boot. hda5 is swap. Grub was allowed to write to the MBR on hda. For both tests I got the cylinder alignment error. For test2, i allowed anaconda to reformat hda6. something changed on both disks that made partition magic complain that the lsb and chs (i think) numbers were inconsistent and it was satified that it could fix them. For test3, partition magic complains that there is something drastically wrong with hdb and tells me not to touch its structure. I have not had any disk related problems actually running either linux or win2k with either release. output from fdisk -l Disk /dev/hda: 40.9 GB, 40982151168 bytes 255 heads, 63 sectors/track, 4982 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start End Blocks Id System /dev/hda1 1 34 273073+ 83 Linux /dev/hda2 35 1309 10241437+ 7 HPFS/NTFS /dev/hda3 * 1310 2202 7173022+ c W95 FAT32 (LBA) /dev/hda4 2203 4982 22330350 f W95 Ext'd (LBA) /dev/hda5 2203 2304 819283+ 82 Linux swap /dev/hda6 2305 4089 14337981 83 Linux /dev/hda7 4090 4982 7172991 b W95 FAT32 Disk /dev/hdb: 81.9 GB, 81964302336 bytes 16 heads, 63 sectors/track, 158816 cylinders Units = cylinders of 1008 * 512 = 516096 bytes Device Boot Start End Blocks Id System /dev/hdb1 * 1 40641 20482843+ 7 HPFS/NTFS Partition 1 does not end on cylinder boundary. /dev/hdb2 40641 81282 20482875 83 Linux /dev/hdb3 81282 97841 8345767+ 7 HPFS/NTFS /dev/hdb4 97841 158802 30724312+ f W95 Ext'd (LBA) /dev/hdb5 97841 158802 30724281 b W95 FAT32 I have the same partition problem after installing FC2. I added a possible fix for it bug 115980 page: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=115980#c33 and more details on the fedora-list-devel: http://www.redhat.com/archives/fedora-test-list/2004-May/msg02114.html Here is a much easier way to fix the partition problem, as suggested by Alexandre Oliva on fedora-test-list: sfdisk -d /dev/hda | sfdisk --no-reread -H240 /dev/hda The 240 value worked for me, in other cases probably 255 is better. Plus, if this helps finding the bug, the kernel still reports wrong disk geometry in /proc/ide/hda/geometry, even after fixing the partition table with the correct values. My machine is a IBM Thinkpad 600e laptop. If I can do anything to help find the problem, please let me know. There is probably a problem with the geometry reported by the kernel. On two machines I installed FC2, this is what /proc/ide/hda/geometry reports: First machine: physical 16383/16/63 logical 19841/16/63 Second machine: physical 16383/16/63 logical 16383/255/63 None of these seem sound (either the physical or logical is wrong as the values do not hold between the two). Can anybody with FC2 confirm this? PS: On a FC1 machine with the 2.4 kernel the numbers are fine: physical 155009/16/63 logical 9726/255/63 One other possible variable - the MBR loader itself. There are several versions of MBR code out there, which behave differently with respect to C/H/S and LBA modes. I added to ms-sys (http://ms-sys.sourceforge.net/) support to write various MBR code in an easy way for exactly this reason - though I was looking at USB boot at the time. ms-sys has been included in the System Rescue CD (http://www.sysresccd.org/). Just try the different MBRs and maybe change partition type to the LBA versions 0c, 0e, 0f. Now parted-1.6.11 is released. Is this version fixed this bug? *** This bug has been marked as a duplicate of 115980 *** Changed to 'CLOSED' state since 'RESOLVED' has been deprecated. |