Currently we are evaluating the GFS-1 SAN on RHEL4.
In our current setup we are having two cluster nodes, attached with two
storage node through Ethernet switch.
On storage nodes we are running the DRBD and heartbeat services.
Every thing is working fine except the following problem in our setup during
during failover testing to check the iscsi we stop the iscsi service on
cluster node. when we restart the service initiator service fails to
reconnect. My understanding is that the iSCSI initiator and iSCSI-target
negotiate the connection parameters through a connection ID which changes
every time we connect an iSCSI initiator service to a target.
do anyone know how we can do this without restarting the cluster node.
Prologix Software Solutions Pvt. Ltd.