Bug 1253833

Summary: renaming logical volumes to previously existing volume names causes confusion when lvresize calls fsadm
Product: Red Hat Enterprise Linux 7 Reporter: Corey Marthaler <cmarthal>
Component: lvm2Assignee: Zdenek Kabelac <zkabelac>
lvm2 sub component: Command-line tools QA Contact: cluster-qe <cluster-qe>
Status: CLOSED ERRATA Docs Contact:
Severity: low    
Priority: unspecified CC: agk, cluster-qe, cmarthal, heinzm, jbrassow, jwaterwo, lmiksik, loberman, lvm-team, mcsontos, msnitzer, prajnoha, prockai, thornber, zkabelac
Version: 7.2   
Target Milestone: rc   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: lvm2-2.02.171-2.el7 Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: 1196910 Environment:
Last Closed: 2017-08-01 21:47:18 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On: 1196910    
Bug Blocks: 1298243, 1385242    

Description Corey Marthaler 2015-08-14 20:17:23 UTC 
+++ This bug was initially created as a clone of Bug #1196910 +++

Description of problem:
renaming logical volumes to previously existing volume names causes confusion when lvresize calls fsadm

Version-Release number of selected component (if applicable):
lvm2-2.02.111-2.el6_6.1.x86_64
device-mapper-1.02.90-2.el6_6.1.x86_64
device-mapper-libs-1.02.90-2.el6_6.1.x86_64

How reproducible:
every time

Steps to Reproduce:
1. mount a volume (volume1)
2. take the volume (volume1) and rename it (volume3)
3. rename another volume (volume2) to the original volume name (volume1)
4. attempt to resize the second volume that has the old name (volume1)

Actual results:
the lvresize command requests that you unmount the mountpoint that is hosting the original volume (volume3)

Expected results:
lvm should be able to determine that this is NOT the same volume and not ask to unmount it


Additional info:
I hit this in a customer's environment, but was able to reproduce on the newest bits in RHEL 6.6

------------------------------
[root@jack-rhel6 ~]# mount /dev/vgtest/volume1 /mnt
[root@jack-rhel6 ~]# cd /mnt
[root@jack-rhel6 mnt]# lvrename vgtest/volume1 volume3
  Renamed "volume1" to "volume3" in volume group "vgtest"
[root@jack-rhel6 mnt]# lvrename vgtest/volume2 volume1
  Renamed "volume2" to "volume1" in volume group "vgtest"
[root@jack-rhel6 mnt]# lvresize -rL512M /dev/vgtest/volume1
Do you want to unmount "/mnt"? [Y|n] n
fsadm: Cannot proceed with mounted filesystem "/mnt"
  fsadm failed: 1
  Filesystem resize failed.
[root@jack-rhel6 mnt]# 
[root@jack-rhel6 mnt]# lvresize -rL512M /dev/vgtest/volume1
Do you want to unmount "/mnt"? [Y|n] y
umount: /mnt: device is busy.
        (In some cases useful info about processes that use
         the device is found by lsof(8) or fuser(1))
fsadm: Cannot proceed with mounted filesystem "/mnt"
  fsadm failed: 1
  Filesystem resize failed.
[root@jack-rhel6 mnt]# 
------------------------------

lvm output shows that volume1 is NOT in open (and not mounted)
------------------------------
[root@jack-rhel6 mnt]# lvs -o +devices
  LV      VG           Attr       LSize Pool Origin Data%  Meta%  Move Log Cpy%Sync Convert Devices        
  lv_root vg_jackrhel6 -wi-ao---- 8.51g                                                     /dev/vda2(0)   
  lv_swap vg_jackrhel6 -wi-ao---- 1.00g                                                     /dev/vda2(2178)
  volume1 vgtest       -wi-a----- 1.00g                                                     /dev/vdb(256)  
  volume3 vgtest       -wi-ao---- 1.00g                                                     /dev/vdb(0)    
[root@jack-rhel6 mnt]# lvresize -rL512M /dev/vgtest/volume1
Do you want to unmount "/mnt"? [Y|n] n
fsadm: Cannot proceed with mounted filesystem "/mnt"
  fsadm failed: 1
  Filesystem resize failed.
[root@jack-rhel6 mnt]# 
[root@jack-rhel6 mnt]# df -h
Filesystem            Size  Used Avail Use% Mounted on
/dev/mapper/vg_jackrhel6-lv_root
                      8.3G  3.3G  4.6G  42% /
tmpfs                 499M   72K  499M   1% /dev/shm
/dev/vda1             477M   75M  377M  17% /boot
/dev/mapper/vgtest-volume1
                     1008M   34M  924M   4% /mnt
------------------------------

Using the resize2fs command instead of passing the -r flag to lvresize works as expected:
------------------------------
[root@jack-rhel6 mnt]# lvresize -rL512M /dev/vgtest/volume1
Do you want to unmount "/mnt"? [Y|n] n
fsadm: Cannot proceed with mounted filesystem "/mnt"
  fsadm failed: 1
  Filesystem resize failed.
[root@jack-rhel6 mnt]# resize2fs /dev/mapper/vgtest-volume1 512M
resize2fs 1.41.12 (17-May-2010)
Resizing the filesystem on /dev/mapper/vgtest-volume1 to 131072 (4k) blocks.
The filesystem on /dev/mapper/vgtest-volume1 is now 131072 blocks long.

[root@jack-rhel6 mnt]# lvresize -rL512M /dev/vgtest/volume1
resize2fs 1.41.12 (17-May-2010)
The filesystem is already 131072 blocks long.  Nothing to do!

  Size of logical volume vgtest/volume1 changed from 1.00 GiB (256 extents) to 512.00 MiB (128 extents).
  Logical volume volume1 successfully resized
[root@jack-rhel6 mnt]# lvresize -rL300M /dev/vgtest/volume1
Do you want to unmount "/mnt"? [Y|n] n
fsadm: Cannot proceed with mounted filesystem "/mnt"
  fsadm failed: 1
  Filesystem resize failed.
[root@jack-rhel6 mnt]#
Comment 1 Corey Marthaler 2015-08-14 20:21:03 UTC 
Exists in 7.2 as well.

3.10.0-302.el7.x86_64
lvm2-2.02.127-1.el7    BUILT: Mon Aug 10 03:22:35 CDT 2015
lvm2-libs-2.02.127-1.el7    BUILT: Mon Aug 10 03:22:35 CDT 2015
lvm2-cluster-2.02.127-1.el7    BUILT: Mon Aug 10 03:22:35 CDT 2015
device-mapper-1.02.104-1.el7    BUILT: Mon Aug 10 03:22:35 CDT 2015
device-mapper-libs-1.02.104-1.el7    BUILT: Mon Aug 10 03:22:35 CDT 2015
device-mapper-event-1.02.104-1.el7    BUILT: Mon Aug 10 03:22:35 CDT 2015
device-mapper-event-libs-1.02.104-1.el7    BUILT: Mon Aug 10 03:22:35 CDT 2015
device-mapper-persistent-data-0.5.4-1.el7    BUILT: Fri Jul 17 08:56:22 CDT 2015
cmirror-2.02.127-1.el7    BUILT: Mon Aug 10 03:22:35 CDT 2015
sanlock-3.2.4-1.el7    BUILT: Fri Jun 19 12:48:49 CDT 2015
sanlock-lib-3.2.4-1.el7    BUILT: Fri Jun 19 12:48:49 CDT 2015
lvm2-lockd-2.02.127-1.el7    BUILT: Mon Aug 10 03:22:35 CDT 2015



[root@host-115 ~]# lvs -a -o +devices
  LV                 Attr       LSize   Cpy%Sync Devices
  resizeA            mwi-aom--- 200.00m 100.00   resizeA_mimage_0(0),resizeA_mimage_1(0)
  [resizeA_mimage_0] iwi-aom--- 200.00m          /dev/sdd2(0)
  [resizeA_mimage_1] iwi-aom--- 200.00m          /dev/sdd1(0)
  resizeB            mwi-aom--- 200.00m 100.00   resizeB_mimage_0(0),resizeB_mimage_1(0)
  [resizeB_mimage_0] iwi-aom--- 200.00m          /dev/sdd2(50)
  [resizeB_mimage_1] iwi-aom--- 200.00m          /dev/sdd1(50)

[root@host-115 ~]# mount
/dev/mapper/mirror_sanity-resizeA on /mnt/resizeA type ext4 (rw,relatime,seclabel,data=ordered)
/dev/mapper/mirror_sanity-resizeB on /mnt/resizeB type ext4 (rw,relatime,seclabel,data=ordered)

[root@host-115 ~]# lvrename /dev/mirror_sanity/resizeA /dev/mirror_sanity/resizeC
  Renamed "resizeA" to "resizeC" in volume group "mirror_sanity"
[root@host-115 ~]# lvrename /dev/mirror_sanity/resizeB /dev/mirror_sanity/resizeA
  Renamed "resizeB" to "resizeA" in volume group "mirror_sanity"

[root@host-115 ~]# lvs -a -o +devices
  LV                 Attr       LSize   Cpy%Sync Devices
  resizeA            mwi-aom--- 200.00m 100.00   resizeA_mimage_0(0),resizeA_mimage_1(0)
  [resizeA_mimage_0] iwi-aom--- 200.00m          /dev/sdd2(50)
  [resizeA_mimage_1] iwi-aom--- 200.00m          /dev/sdd1(50)
  resizeC            mwi-aom--- 200.00m 100.00   resizeC_mimage_0(0),resizeC_mimage_1(0)
  [resizeC_mimage_0] iwi-aom--- 200.00m          /dev/sdd2(0)
  [resizeC_mimage_1] iwi-aom--- 200.00m          /dev/sdd1(0)

[root@host-115 ~]# lvextend -r -n -f -L 400M /dev/mirror_sanity/resizeA
  Extending 2 mirror images.
  Size of logical volume mirror_sanity/resizeA changed from 200.00 MiB (50 extents) to 400.00 MiB (100 extents).
  Logical volume resizeA successfully resized
resize2fs 1.42.9 (28-Dec-2013)
resize2fs: Device or resource busy while trying to open /dev/mapper/mirror_sanity-resizeA
Couldn't find valid filesystem superblock.
fsadm: Resize ext4 failed
  fsadm failed: 1

[root@host-115 ~]# lvs -a -o +devices
  LV                 Attr       LSize   Cpy%Sync Devices
  resizeA            mwi-aom--- 400.00m 100.00   resizeA_mimage_0(0),resizeA_mimage_1(0)
  [resizeA_mimage_0] iwi-aom--- 400.00m          /dev/sdd2(50)
  [resizeA_mimage_1] iwi-aom--- 400.00m          /dev/sdd1(50)
  resizeC            mwi-aom--- 200.00m 100.00   resizeC_mimage_0(0),resizeC_mimage_1(0)
  [resizeC_mimage_0] iwi-aom--- 200.00m          /dev/sdd2(0)
  [resizeC_mimage_1] iwi-aom--- 200.00m          /dev/sdd1(0)
Comment 2 Zdenek Kabelac 2015-08-24 08:33:21 UTC 
Yep - this will need some more tricky behaviour from fsadm size, since renamed LVs are unfortunately remembered under the old name in kernel.
Comment 7 Zdenek Kabelac 2017-05-09 11:14:52 UTC 
So this bug is getting a bit more complicated.

There are couple connected troubles.

fsadm needs to provide better detection of mounted volume - as it appears most tools do a very basic detect from  /etc/mtab content which keeps the old name.

New systems may use  /proc/self/mountinfo - but it's not giving all the needed info either (i.e. cgroups)  but the advantage here is the we get  major:minor.

We can get a similar info in somewhat complicated way also via 'stat'.


So the fix will try to address couple issues  accross mutliple different tools.

I'm aiming ATM to prevent and  filesystem damage by disallowing overrides for cases were  fsadm failed to recognize mounted filesystem.

But few more related bugs will need to be opened against tools: resize2fs, reiser_resizefs,  xfs_growfs  - which needs to provide smarter logic to handle renamed device better.
Comment 8 Alasdair Kergon 2017-05-11 23:04:52 UTC 
So the approach for 7.4 is to update fsdam with additional checks to detect situations that the current code handles incorrectly and provide clear error messages.

Beyond that, wider changes will be covered by separate bugzillas.
Comment 9 Zdenek Kabelac 2017-05-24 15:56:41 UTC 
Fixed by several upstream fsadm commits:

https://www.redhat.com/archives/lvm-devel/2017-May/msg00061.html

this one was the most important - ensuring lvm2 will not use renamed mounted LV for filesystem utils.
Comment 10 Corey Marthaler 2017-05-25 20:15:23 UTC 
Is this the fix for this scenario? This behavior is now in the latest rpms and causing this case to "fail".


lvreduce -y -r -n -f -L 500M /dev/snapper_thinp/other1
fsadm: Cannot RESIZE device "/dev/mapper/snapper_thinp-other1" without umounting filesystem /mnt/other2 first.
fsadm: Mounted filesystem is using device snapper_thinp-other1, but referenced device is snapper_thinp-other2.
fsadm: Filesystem utilities currently do not support renamed devices.
  /usr/sbin/fsadm failed: 1
  Filesystem resize failed.
couldn't resize thin and filesystem on host-126


3.10.0-666.el7.x86_64

lvm2-2.02.171-2.el7    BUILT: Wed May 24 09:02:34 CDT 2017
lvm2-libs-2.02.171-2.el7    BUILT: Wed May 24 09:02:34 CDT 2017
lvm2-cluster-2.02.171-2.el7    BUILT: Wed May 24 09:02:34 CDT 2017
device-mapper-1.02.140-2.el7    BUILT: Wed May 24 09:02:34 CDT 2017
device-mapper-libs-1.02.140-2.el7    BUILT: Wed May 24 09:02:34 CDT 2017
device-mapper-event-1.02.140-2.el7    BUILT: Wed May 24 09:02:34 CDT 2017
device-mapper-event-libs-1.02.140-2.el7    BUILT: Wed May 24 09:02:34 CDT 2017
device-mapper-persistent-data-0.7.0-0.1.rc6.el7    BUILT: Mon Mar 27 10:15:46 CDT 2017
Comment 11 Zdenek Kabelac 2017-05-26 08:40:14 UTC 
Yes - that's currently intended behavior -  since  filesystem utils needs quite some fixes to be able to handle this case.

I'll attach created individual BZ here for tracking.

Once utils are fixed - lvm2 will need to detect proper/fixed version of tools being present in system and let them proceed with operation in this case.
Comment 13 Corey Marthaler 2017-06-01 15:29:31 UTC 
Marking verified in the latest rpms. Based on comment #11, checked that these reduce/extend operations are no longer allowed until the devices have been remounted.

3.10.0-672.el7.x86_64
lvm2-2.02.171-3.el7    BUILT: Wed May 31 08:36:29 CDT 2017
lvm2-libs-2.02.171-3.el7    BUILT: Wed May 31 08:36:29 CDT 2017
lvm2-cluster-2.02.171-3.el7    BUILT: Wed May 31 08:36:29 CDT 2017
device-mapper-1.02.140-3.el7    BUILT: Wed May 31 08:36:29 CDT 2017
device-mapper-libs-1.02.140-3.el7    BUILT: Wed May 31 08:36:29 CDT 2017
device-mapper-event-1.02.140-3.el7    BUILT: Wed May 31 08:36:29 CDT 2017
device-mapper-event-libs-1.02.140-3.el7    BUILT: Wed May 31 08:36:29 CDT 2017
device-mapper-persistent-data-0.7.0-0.1.rc6.el7    BUILT: Mon Mar 27 10:15:46 CDT 2017



# extend
lvextend -r -n -f -L 1.5G /dev/snapper_thinp/other1
  WARNING: Sum of all thin volume sizes (6.50 GiB) exceeds the size of thin pool snapper_thinp/POOL (1.00 GiB)!
fsadm: Cannot RESIZE device "/dev/mapper/snapper_thinp-other1" without umounting filesystem /mnt/other2 first.
fsadm: Mounted filesystem is using device snapper_thinp-other1, but referenced device is snapper_thinp-other2.
fsadm: Filesystem utilities currently do not support renamed devices.
  /usr/sbin/fsadm failed: 1


# reduce
lvreduce -y -r -n -f -L 500M /dev/snapper_thinp/other1
fsadm: Cannot RESIZE device "/dev/mapper/snapper_thinp-other1" without umounting filesystem /mnt/other2 first.
fsadm: Mounted filesystem is using device snapper_thinp-other1, but referenced device is snapper_thinp-other2.
fsadm: Filesystem utilities currently do not support renamed devices.
  /usr/sbin/fsadm failed: 1
  Filesystem resize failed.
Comment 14 errata-xmlrpc 2017-08-01 21:47:18 UTC 
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2017:2222