[Gluster-users] Cannot stop (and delete) "faulty" geo-replication because slave is gone

[Fred van Zwieten]

Hello,

I am trying to setup (as a test) geo-replication between a KVM Fedora17
guest with gluster 3.3.1 on it and an Amazon EC2 instance running fedora17.

The slave (EC2 instance) is gone and I want to remove the geo-replication
relation. It's status is, naturally, "faulty". It refused to delete  with
the following message:

[root at g61 ~]# gluster volume geo-replication vol1
ssh://georepl@g81.lab8.local::cloudvol1 stop
geo-replication session between vol1 & ssh://georepl@g81.lab8.local::cloudvol1
not active
geo-replication command failed

I know it's not active, because the slave is gone. So naturally I cannot
start it either:

[root at g61 ~]# gluster volume geo-replication vol1
ssh://georepl@g81.lab8.local::cloudvol1 start
geo-replication session between vol1 & ssh://georepl@g81.lab8.local::cloudvol1
already started
geo-replication command failed

Now what?

Greets,

Fred
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20121202/f354f0b4/attachment.html>