[Gluster-users] xattrs errors when creating files on Gluster3.3b2

Daniel Pereira d.pereira at skillupjapan.co.jp
Thu Mar 22 03:21:23 UTC 2012

Hi everyone,

Hope anyone can shed some light on some strange behavior we've just 
having with Gluster.
Last night, while using gluster3.3b2, we came up with some errors, after 
running tests for several days.

We're running the test in a striped 4 disk volume:
# gluster volume info data
Volume Name: data
Type: Stripe
Status: Started
Number of Bricks: 4
Transport-type: tcp

When creating a lot of files, some of the files started to have erratic 
behavior. They are in the file system (ls commands below) but we are not 
being able to read some of the files (file ending with 22413, in the 
command below), 9 in total to be exact.
Both files were created programatically, under the same circumstances at 
nearly the same time. The directory has currently 12668 files inside.

# ls -la fd66e046-6c2f-11e1-8e8d-003048bfe717-segment22413-22413.mp4
-rw-r--r--. 1 1000 1000 41759951 Mar 21 20:34 
# ls -la fd66e046-6c2f-11e1-8e8d-003048bfe717-segment22412-22412.mp4
-rw-r--r--. 1 1000 1000 40994275 Mar 21 20:34 

# file fd66e046-6c2f-11e1-8e8d-003048bfe717-segment22413-22413.mp4
fd66e046-6c2f-11e1-8e8d-003048bfe717-segment22413-22413.mp4: writable, 
regular file, no read permission
# file fd66e046-6c2f-11e1-8e8d-003048bfe717-segment22412-22412.mp4
fd66e046-6c2f-11e1-8e8d-003048bfe717-segment22412-22412.mp4: ISO Media, 
MPEG v4 system, 3GPP JVT AVC

When we checked the logs of each brick, all of the bricks had these 
errors for the 9 files in question:

[2012-03-21 20:34:36.309525] E [posix-helpers.c:585:posix_handle_pair] 
key:trusted.data-stripe-0.stripe-size error:File exists
gluster-enclusterdata_4.log:[2012-03-21 20:34:36.309580] E 
[posix.c:1766:posix_create] 0-data-posix: setting xattrs on 
failed (File exists)
gluster-enclusterdata_4.log:[2012-03-22 08:15:09.205467] I 
[server3_1-fops.c:1493:server_stat_cbk] 0-data-server: 176411: STAT 
(0) ==> -1 (No such file or directory)
gluster-enclusterdata_4.log:[2012-03-22 08:15:29.287542] I 
[server3_1-fops.c:1493:server_stat_cbk] 0-data-server: 184724: STAT 
(0) ==> -1 (No such file or directory)

Whenever we try to read the file, the "No such file or directory" error 
crops up in the logs.
Any idea what could be going wrong? In total we have 9 files that show 
this erratic behavior, all with similar information on the brick logs.

Thanks in advance,

More information about the Gluster-users mailing list