1. Create a disk
Disk initialization, creating volumes, and formatting.
2. Create two folders on the creation disk
One is used to map to an arbitration disk, and the other is used to make a disk.
3. Create a role:
Adding iSCSI Services requires a file server with VDS VSS as well as the iSCSI target service.
4. Create an iSCSI virtual disk
Using a custom path to create, select the folder you just created.
Adding an access target is to add a client server that needs to be accessed.
Name, which is the target name of the originating iSCSI program.
There is also a virtual disk name that does not affect the use.
5. iSCSI Program Initiation
In the administration tool, there is an iSCSI program initiator.
Each server, through the discovery of the iSCSI program initiator, adds the IP address of the target iSCSI server, then on the Target tab page, then connects, and then on the Devices tab, automatically configures.
This configuration is required for each server
6. Disk Initialization
Using one of the servers, through the management tool, the disk initialization and mount.
The other two also need to mount the disk and bring it online.
7. Establishing a Failover Cluster
The Failover Clustering feature is installed on each server.
The cluster is then created on top of the machine as the master node.
8, the role of HA
The main ha is to do dual-machine hot standby. MSCS also supports multiple devices joining the same failover cluster, but there are only 1 external services available at the same time, so it makes no sense to add more machines.
If you want a multi-machine service, you need to do something like NLB.
9. The limitation of MSCS is that, above the failover cluster, roles need to be added for service. So you need to support the role of Windows to do ha, for the unsupported can not do.
10. Recovery role, fail back role
If a fails, if a recovery fails, whether or not to preempt changes back to the host.
If preemption, it is necessary to interrupt, if not preemption, it is not required.
"Windows system Technology" MSCS failover cluster