[Gluster-devel] Quota Used Value Incorrect - Fix now or after upgrade

[Matthew B]

Hello,

I need some advice on fixing an issue with quota on my gluster volume. It's
running version 3.7, distributed volume, with 7 nodes.

# gluster --version
glusterfs 3.7.13 built on Jul  8 2016 15:26:18
Repository revision: git://git.gluster.com/glusterfs.git
Copyright (c) 2006-2011 Gluster Inc. <http://www.gluster.com>
GlusterFS comes with ABSOLUTELY NO WARRANTY.
You may redistribute copies of GlusterFS under the terms of the GNU General
Public License.

# gluster volume info storage

Volume Name: storage
Type: Distribute
Volume ID: 6f95525a-94d7-4174-bac4-e1a18fe010a2
Status: Started
Number of Bricks: 7
Transport-type: tcp
Bricks:
Brick1: 10.0.231.50:/mnt/raid6-storage/storage
Brick2: 10.0.231.51:/mnt/raid6-storage/storage
Brick3: 10.0.231.52:/mnt/raid6-storage/storage
Brick4: 10.0.231.53:/mnt/raid6-storage/storage
Brick5: 10.0.231.54:/mnt/raid6-storage/storage
Brick6: 10.0.231.55:/mnt/raid6-storage/storage
Brick7: 10.0.231.56:/mnt/raid6-storage/storage
Options Reconfigured:
changelog.changelog: on
geo-replication.ignore-pid-check: on
geo-replication.indexing: on
nfs.disable: no
performance.readdir-ahead: on
features.quota: on
features.inode-quota: on
features.quota-deem-statfs: on
features.read-only: off

# df -h /storage/
Filesystem            Size  Used Avail Use% Mounted on
10.0.231.50:/storage  255T  172T   83T  68% /storage


I am planning to upgrade to 3.10 (or 3.12 when it's available) but I have a
number of quotas configured, and one of them (below) has a very wrong
"Used" value:

# gluster volume quota storage list | egrep "MEOPAR "
/data/projects/MEOPAR                      8.5TB     80%(6.8TB) *16384.0PB*
17.4TB              No                   No


I have confirmed the bad value appears in one of the bricks current xattrs,
and it looks like the issue has been encountered previously on bricks 04,
03, and 06: (gluster07 does not have a trusted.glusterfs.quota.size.1 as it
was recently added)

$ ansible -i hosts gluster-servers[0:6] -u <user> --ask-pass -m shell
-b --become-method=sudo --ask-become-pass -a "getfattr
--absolute-names -m . -d -e hex
/mnt/raid6-storage/storage/data/projects/MEOPAR | egrep
'^trusted.glusterfs.quota.size'"
SSH password:
SUDO password[defaults to SSH password]:

gluster02 | SUCCESS | rc=0 >>
trusted.glusterfs.quota.size=0x0000011ecfa56c00000000000005cd6d000000000006d478
trusted.glusterfs.quota.size.1=0x0000010ad4a452000000000000012a0300000000000150fa

gluster05 | SUCCESS | rc=0 >>
trusted.glusterfs.quota.size=0x00000033b8e92200000000000004cde8000000000006b1a4
trusted.glusterfs.quota.size.1=0x0000010dca277c00000000000001297d0000000000015005

gluster01 | SUCCESS | rc=0 >>
trusted.glusterfs.quota.size=0x0000003d4d4348000000000000057616000000000006afd2
trusted.glusterfs.quota.size.1=0x00000133fe211e00000000000005d161000000000006cfd4

gluster04 | SUCCESS | rc=0 >>
trusted.glusterfs.quota.size=0xffffff396f3e9400000000000004d7ea0000000000068c62
trusted.glusterfs.quota.size.1=0x00000106e6724800000000000001138f0000000000012fb2

gluster03 | SUCCESS | rc=0 >>
trusted.glusterfs.quota.size=0xfffffd02acabf000000000000003599000000000000643e2
trusted.glusterfs.quota.size.1=0x00000114e20f5e0000000000000113b30000000000012fb2

gluster06 | SUCCESS | rc=0 >>
trusted.glusterfs.quota.size=0xffffff0c98de440000000000000536e40000000000068cf2
trusted.glusterfs.quota.size.1=0x0000013532664e00000000000005e73f000000000006cfd4

gluster07 | SUCCESS | rc=0 >>
*trusted.glusterfs.quota.size=0xfffffa3d7c1ba60000000000000a9ccb000000000005fd2f*

And reviewing the subdirectories of that folder on the impacted server you
can see that none of the direct children have such incorrect values:

[root at gluster07 ~]# getfattr --absolute-names -m . -d -e hex
/mnt/raid6-storage/storage/data/projects/MEOPAR/*
# file: /mnt/raid6-storage/storage/data/projects/MEOPAR/<dir1>
...
trusted.glusterfs.quota.7209b677-f4b9-4d82-a382-0733620e6929.contri=0x000000fb6841820000000000000074730000000000000dae
trusted.glusterfs.quota.dirty=0x3000
trusted.glusterfs.quota.size=0x000000fb6841820000000000000074730000000000000dae

# file: /mnt/raid6-storage/storage/data/projects/MEOPAR/<dir2>
...
trusted.glusterfs.quota.7209b677-f4b9-4d82-a382-0733620e6929.contri=0x0000000416d5f4000000000000000baa0000000000000441
trusted.glusterfs.quota.dirty=0x3000
trusted.glusterfs.quota.limit-set=0x0000010000000000ffffffffffffffff
trusted.glusterfs.quota.size=0x0000000416d5f4000000000000000baa0000000000000441

# file: /mnt/raid6-storage/storage/data/projects/MEOPAR/<dir3>
...
trusted.glusterfs.quota.7209b677-f4b9-4d82-a382-0733620e6929.contri=0x000000110f2c4e00000000000002a76a000000000006ad7d
trusted.glusterfs.quota.dirty=0x3000
trusted.glusterfs.quota.limit-set=0x0000020000000000ffffffffffffffff
trusted.glusterfs.quota.size=0x000000110f2c4e00000000000002a76a000000000006ad7d


Can I fix this on the current version of gluster (3.7) on just the one
brick before I upgrade? Or would I be better off upgrading to 3.10 and
trying to fix it then?

I have reviewed information here:

http://lists.gluster.org/pipermail/gluster-devel/2016-February/048282.html
http://lists.gluster.org/pipermail/gluster-users.old/2016-September/028365.html

It seems like since I am on Gluster 3.7 I can disable quotas and re-enable
and everything will get recalculated and increment the index on the
quota.size xattr. But with the size of the volume that will take a very
long time.

Could I simply mark the impacted directly as dirty on gluster07? Or update
the xattr directly as the sum of the size of dir1, 2, and 3?

Thanks,
 -Matthew
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gluster.org/pipermail/gluster-devel/attachments/20170825/dee96a2b/attachment-0001.html>