Gluster users - Dec 2014 - Is it possible to start geo-replication between two volumes with data already present in the slave?

Hello again,

Ok - I managed to get a list of gfids and pathnames transferred to the slave
server.

I began the process of updating the heal - but I get a bunch of stale file
handles when I try and run the following on the slave volume:

                    cd $GLUSTER_SRC/extras/geo-rep
        sh slave-upgrade.sh localhost:<SLAVE VOL NAME>
/tmp/master-gfid-values.txt $PWD/gsync-sync-gfid

setxattr on ./serverTempDir/FOO failed (Stale file handle)
...

This maps to a [fuse-bridge.c:1261:fust_err_cbk] 0-glusterfs-fuse error in
/var/log/glusterfs/tmp-mount-log

Any thoughts on this?

Thanks in advance,

Nathan


From: Nathan Aldridge
Sent: Monday, December 15, 2014 11:55 AM
To: 'gluster-users at gluster.org'
Subject: RE: [Gluster-users] Is it possible to start geo-replication between two
volumes with data already present in the slave?

Hi,

Thanks for your response. I attempted to run the
"generate-gfid-file.sh" script but it failed when it encountered
"whitespace" in some of the filenames. I'll take a look to see if
I can see what is breaking in the scripts... but if you know offhand, that would
be appreciated? I see a whole lot of getfattr errors due to not being able to
find a file.

Cheers,

Nathan Aldridge
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://supercolony.gluster.org/pipermail/gluster-users/attachments/20141217/fff00c90/attachment.html>