Hello,
I have put together a list of tasks for Geo-Replication based of our offcial documentation on Red Hat Gluster Storage :
This is what I found:
Gluster 3.1:
Step by Step Geo-Replication:
1- Enable shared storage on master node
#gluster volume set all cluster.enable-shared-storage enable
2 - create a SSH keypair without a passphrase on master node
#ssh-keygen
3 - Copy the SSH key password less SSH access to your slave nodes (on master node)
#ssh-copy-id <standard user>@<slave node>
4 - On the slave node
#mkdir /var/mountbroker-root
#chmod 0711 /var/mountbroker-root
#semanage fcontext -a -e /home /var/mountbroker-root
#restorecon -Rv /var/mountbroker-root
5 - Configure the options for the glusterd and restart service (on slave node)
#gluster system:: execute mountbroker opt mountbroker-root /var/mountbroker-root
#gluster system:: execute mountbroker user geoaccount slavevol
#gluster system:: execute mountbroker opt geo-replication-log-group geogroup
#gluster system:: execute mountbroker opt rpc-auth-allow-insecure on
# systemctl restart glusterd
6 - On the master node
#gluster system:: execute gsec_create
#gluster volume geo-replication mastervol geoaccount@servere::slavevol create push-pem
7 - On slave server
#/usr/libexec/glusterfs/set_geo_rep_pem_keys.sh geoaccount mastervol slavevol
8 - On Master
#gluster volume geo-replication mastervol geoaccount@servere::slavevol config use_meta_volume true
#gluster volume geo-replication mastervol geoaccount@servere::slavevol start
Gluster 3.2:
On the slave node:
#gluster-mountbroker setup </var/mountbroker-root> <geogroup>
#gluster-mountbroker add <slavevol> <geoaccount>
#systemctl restart glusterd
On the Master node:
#ssh-keygen
#ssh-copy-id <geoaccount>@<slave_node>
#gluster system:: execute gsec_create
#gluster volume geo-replication <mastervol> <geoaccount>@<slave_node>::<slavevol> create push-pem
On the Slave node:
/usr/libexec/glusterfs/set_geo_rep_pem_keys.sh <geoaccount> <mastervol> <slavevol>
On the Master node:
#gluster volume geo-replication <mastervol> <slave_node>::<slavevol> config use_meta_volume true
#gluster volume geo-replication <mastervol> <geoaccount>@<slave_node>::<slavevol> start
#gluster volume geo-replication <mastervol> <geoaccount>@<slave_node>::<slavevol> status
You can see that on Gluster 3.2 we don't have to type as much, but my questions are:
3.1:
#mkdir /var/mountbroker-root
#chmod 0711 /var/mountbroker-root
#semanage fcontext -a -e /home /var/mountbroker-root
#restorecon -Rv /var/mountbroker-root
#gluster system:: execute mountbroker opt mountbroker-root /var/mountbroker-root
#gluster system:: execute mountbroker user geoaccount slavevol
#gluster system:: execute mountbroker opt geo-replication-log-group geogroup
#gluster system:: execute mountbroker opt rpc-auth-allow-insecure on
# systemctl restart glusterd
3.2:
#gluster-mountbroker setup </var/mountbroker-root> <geogroup>
#gluster-mountbroker add <slavevol> <geoaccount>
#systemctl restart glusterd
Are those two steps equivalent (given the correct Gluster version) ?
Thank you,