[Gluster-users] Rsync error when brick become online after a reboot / cash

[Xavier Normand]

I'm currently doing some HA testing on glusterfs and i notice a problem
when a cluster node become online after a reboot or a crash. There is my
setup:

 Volume Name: data
Type: Replicate
Status: Started
Number of Bricks: 2
Transport-type: tcp
Bricks:
Brick1: gluster1:/data
Brick2: gluster2:/data

 Before starting rsync, i make sure glusterd is closed on brick1. I start
the rsync " about 160 MB " and right after the rsync begin, i start
glusterd on brick1. After a couple seconds, rsync show "rsync: rename
"/mount/file/path/file.sh.sMZV6w" -> "/source/file/path/file.sh": No such
file or directory (2)"

 On brick1 gluster log : [2012-03-07 10:18:22.82195] I
[server3_1-fops.c:259:server_inodelk_cbk] 0-data-server: 907653: INODELK
/local/brick/path/file/fie.sh (0) ==> -1 (No such file or directory)

 If i check in the gluster mount the file isn't there!

 Any clue?
*Xavier Normand*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20120308/b2635d58/attachment.html>