[Gluster-users] Moved files from one directory to another, now gone

Andrus, Brian Contractor bdandrus at nps.edu
Thu Mar 3 19:50:09 UTC 2016


Pranith,

Sorry for taking so long. Work required my time.

So I can look at this a little more right now.
I am using glusterfs-3.7.6-1
Here is some pertinent info:
]# gluster volume info DATA

Volume Name: DATA
Type: Disperse
Volume ID: 8374314b-796a-48d0-acf7-dea51fdd83ae
Status: Started
Number of Bricks: 1 x (2 + 1) = 3
Transport-type: tcp
Bricks:
Brick1: server45:/zpool1/DATABRICK
Brick2: server46:/zpool1/DATABRICK
Brick3: server47:/zpool1/DATABRICK
Options Reconfigured:
performance.io-cache: off
performance.quick-read: off
performance.readdir-ahead: disable

# df -h /zpool1 /DATA
Filesystem      Size  Used Avail Use% Mounted on
zpool1          7.2T  1.2T  6.0T  17% /zpool1
server45:/DATA   15T  2.5T   12T  18% /DATA

# du -hsc /DATA/
232G    /DATA/
232G    total

I have been cleaning up some of the stuff, but there are 2 directories I cannot remove:
# ls /DATA/
FINISHED_LOGS  home  JOBS  LOGS  m4p2  OLDLOGS

# ls /zpool1/DATABRICK/
FINISHED_LOGS/ .glusterfs/    home/          JOBS/          LOGS/          m4p2/          OLDLOGS/       .trashcan/

# ls /zpool1/DATABRICK/
FINISHED_LOGS  home  JOBS  LOGS  m4p2  OLDLOGS

# rm -rf /DATA/OLDLOGS/
# rm -rf /DATA/FINISHED_LOGS/
# ls /DATA/
FINISHED_LOGS  home  JOBS  LOGS  m4p2  OLDLOGS

Those two directories are empty on the glusterfs mount, but still have files when I look on the brick.

Strange bug. I will be just reformatting the thing when some of the work completes, but I can run some tests and provide info if you want to try and track it down.

Brian Andrus



From: Pranith Kumar Karampuri [mailto:pkarampu at redhat.com]
Sent: Thursday, January 28, 2016 10:36 PM
To: Andrus, Brian Contractor <bdandrus at nps.edu>; gluster-users at gluster.org
Subject: Re: [Gluster-users] Moved files from one directory to another, now gone


On 01/29/2016 12:50 AM, Andrus, Brian Contractor wrote:
All,

I have a glusterfs setup with a disperse volume over 3 zfs bricks, one on each node.

I just did a 'mv' of some log files from one directory to another and when I look in the directory, they are not there at all!
Neither is any of the data I used to have. It is completely empty. I try doing a 'touch' of a file and nothing shows up.

I do see the files on the zpool mount, but they seem to be partially there. They are all just text files, but they look incomplete.
There seems to be many errors in the log:

[2016-01-28 18:57:41.353265] W [MSGID: 122002] [ec-common.c:71:ec_heal_report] 0-DATA-disperse-0: Heal failed [Transport endpoint is not connected]

Any ideas on what this is? Other directories seem ok.

Could you give me the following information:
1) Output of: gluster volume info
2) Log files of the setup? Both client and brick logs, they should be present in /var/log/glusterfs
3) What is the version of glusterfs you are using?

Pranith


Thanks in advance,
Brian Andrus





_______________________________________________

Gluster-users mailing list

Gluster-users at gluster.org<mailto:Gluster-users at gluster.org>

http://www.gluster.org/mailman/listinfo/gluster-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.gluster.org/pipermail/gluster-users/attachments/20160303/1d7b91ee/attachment.html>


More information about the Gluster-users mailing list