Recently, I encountered this problem when resizing an EXT4 file system in our NFS server.
# df -hT /export/rear
Filesystem Type Size Used Avail Use% Mounted on
/dev/mapper/vgadm-lvrear
ext4 1.2T 1.2T 16M 100% /export/rear
# resize2fs /dev/mapper/vgadm-lvrear
resize2fs 1.41.12 (17-May-2010)
Filesystem at /dev/mapper/vgadm-lvrear is mounted on /export/rear; on-line resizing required
old desc_blocks = 74, new_desc_blocks = 108
Performing an on-line resize of /dev/mapper/vgadm-lvrear to 450920448 (4k) blocks.
resize2fs: Operation not permitted While trying to add group #9472
# dmesg
EXT4-fs warning (device dm-28): ext4_group_add: No reserved GDT blocks, can't resize
EXT4-fs warning (device dm-28): ext4_group_add: No reserved GDT blocks, can't resize
I followed this Redhat KB Article - https://access.redhat.com/solutions/88483
# tune2fs -O ^has_journal /dev/mapper/vgadm-lvrear
tune2fs 1.41.12 (17-May-2010)
The has_journal feature may only be cleared when the filesystem is
unmounted or mounted read-only.
# dumpe2fs /dev/mapper/vgadm-lvrear | grep -i features
dumpe2fs 1.41.12 (17-May-2010)
Filesystem features: has_journal ext_attr resize_inode dir_index filetype needs_recovery extent flex_bg sparse_super large_file huge_file uninit_bg dir_nlink extra_isize
Journal features: journal_incompat_revoke
So solution would be either:
- Go and install XFS (need additional subscription as the system is RHEL 6.8)
- Or recreate the file system but with additional parameters --resize=max-online-resize
Full explanation has been posted here as well.
Thus, it has been decided by our Platform Architect that it's okay to move to XFS instead. So I needed to enable a subscription and went ahead installing XFS.
$ sudo subscription-manager repos --list | grep scalefs
$ sudo subscription-manager repos --enable=rhel-6.4-scalefs-server-for-rhs-2.1-rpms
Here's what I did.
$ sudo yum install xfsprogs xfsdump -y
I made a request of 2x1TB LUNs from our Storage Team and created a new volume group called vgrear.
$ sudo pvcreate /dev/mapper/3600601604bc03800c862f9bc0bdce711
$ sudo pvcreate /dev/mapper/3600601604bc03800c662f9bc0bdce711
$ sudo vgcreate vgrear /dev/mapper/3600601604bc03800c862f9bc0bdce711 /dev/mapper/3600601604bc03800c662f9bc0bdce711
Created the logical volume group and the new file system (/export/rear_new).
$ sudo mkfs.xfs /dev/vgrear/lvrear
meta-data=/dev/vgrear/lvrear isize=256 agcount=4, agsize=134217216 blks
= sectsz=512 attr=2, projid32bit=0
data = bsize=4096 blocks=536868864, imaxpct=5
= sunit=0 swidth=0 blks
naming =version 2 bsize=4096 ascii-ci=0
log =internal log bsize=4096 blocks=262143, version=2
= sectsz=512 sunit=0 blks, lazy-count=1
realtime =none extsz=4096 blocks=0, rtextents=0
$ sudo mkdir /export/rear_new
$ sudo mount /dev/vgrear/lvrear /export/rear_new
$ df -hT /export/rear_new
Filesystem Type Size Used Avail Use% Mounted on
/dev/mapper/vgrear-lvrear
xfs 2.0T 33M 2.0T 1% /export/rear_new
Afterwards, I did an rsync (copy/transfer of files) from /export/rear to /export/rear_new. And did the un/export, un/re-mounting of the file systems (did an lsof to ensure nothing is using the files/folders). I also modified the /etc/fstab to reflect the changes I made in the VG/LV.
$ sudo rsync -avP /export/rear/* /export/rear_new/
sending incremental file list
a.out
...
$ sudo lsof /export/rear_new/
$ sudo lsof /export/rear
$ sudo exportfs -u *:/export/rear
$ sudo umount /export/rear; sudo umount /export/rear_new
$ sudo vi /etc/fstab
$ cat /etc/fstab | grep rear
/dev/vgadm/lvrear /export/rear_old ext4 ro,data=ordered 0 0
/dev/vgrear/lvrear /export/rear xfs defaults,nodev 0 0
$ sudo mv /export/rear /export/rear_old
$ sudo mv /export/rear_new /export/rear
$ ls -l /export/ | grep rear
drwxrwxrwx 220 root root 8192 Dec 13 18:56 rear
drwxrwxrwx 221 root root 12288 Dec 13 06:23 rear_old
$ sudo mount /dev/vgrear/lvrear /export/rear
$ sudo mount /dev/vgadm/lvrear /export/rear_old
$ df -hT | grep rear
/dev/mapper/vgrear-lvrear
xfs 2.0T 1021G 1.1T 50% /export/rear
/dev/mapper/vgadm-lvrear
ext4 1.2T 1.0T 143G 88% /export/rear_old
And tested the archived file system /export/rear_old, and now it is read-only mode.
$ sudo touch /export/rear_old/test.ro
touch: cannot touch `/export/rear_old/test.ro': Read-only file system
