Windows Server 2008 End-to-End Storage Scenarios
1. Windows Server 2008 Storage Explorer
1A. Goal
Making sure the devices are properly discovered.
1B1. Setup for FC
FC Network: Make sure hosts, switches and SAN are properly connected.
SAN: Configure device (including configuration of WWNs), create/unmask LUNs.
Host: Load/configure HBA drivers.
1B2. Setup for iSCSI
Network: Make sure hosts, switches and SAN are properly connected.
iSNS: Install iSNS Server on the network.
SAN: Configure iSCSI device (including iSNS and targets), create/unmask LUNs.
Host: Load/configure network drivers. Configure iSCSI initiator (including iSNS).
1C. Test
In Storage Explorer, verify that all devices/LUNs are properly discovered.
In Storage Explorer, verify that you launch the device’s native management tool.
Unplug cable connecting switch to the SAN, then refresh the view In Storage Explorer to verify that information is properly updated.
2. Windows Server 2008 Storage Manager for SANS (SMfS)
2A. Goal
Testing the devices Virtual Disk Services (VDS) Hardware Providers.
2B. Setup
Network: Make sure hosts, switches and SAN are properly connected.
SAN: Configure device.
Host: Load/configure HBA drivers or iSCSI Initiator. Load VDS Hardware Provider. In Server Manager, add the Storage Manager for SANs feature.
2C. Test
In Storage Manager for SANS (SMfS), verify the storage subsystem has been discovered.
In Storage Manager for SANS (SMfS), create a LUN, assign it to the server and create a volume on it.
3. Windows Server 2008 Extend and Shrink
3A. Goal
Create a volume, extend the volume, then shrink a volume.
3B. Setup
Network: Make sure hosts, switches and SAN are properly connected.
SAN: Configure device.
Host: Load/configure HBA drivers or iSCSI Initiator. Load VDS Hardware Provider. In Server Manager, add the Storage Manager for SANs feature.
3C. Test
In Storage Manager for SANS (SMfS), create a LUN with 300 GB, assign it to the server and create a volume on it.
Copy files to the volume.
In Storage Manager for SANS (SMfS), extend the LUN to 600 GB.
In Disk Management, extend the volume to 600 GB.
Copy more files to the volume.
Delete files create before extending the volume.
In Disk Management, shrink the volume back to 300 GB.
Use DISKRAID.EXE to shrink the LUN back to 300 GB.
3D. Additional Information
DISKRAID.EXE commands to shrink a LUN:
SELECT SUBSYSTEM
LIST LUNS
SELECT LUN
SHRINK LUN
LIST LUNS
4. Windows Server 2008 Volume Shadowcopy Services (VSS)
4A. Goal
Use VSS Hardware Provider to create a snapshot.
4B. Setup
Network: Make sure hosts, switches and SAN are properly connected.
SAN: Configure device.
Host: Load/configure HBA drivers or iSCSI Initiator. Load VDS Hardware Provider. In Server Manager, add the Storage Manager for SANs feature. Load VSS Hardware Provider.
4C. Test
In Storage Manager for SANS (SMfS), create a LUN, assign it to the server and create a volume on it.
Copy files to volume.
Create a hardware snapshot of the volume using DISKSHADOW.EXE.
Delete files from volume.
Expose the snapshot of the volume using DISKSHADOW.EXE
4D. Additional information
DISKSHADOW.EXE commands to create a snapshot:
LIST PROVIDERS
SET CONTEXT PERSISTENT
ADD VOLUME E:
CREATEDISKSHADOW.EXE commands to expose a snapshot:
LIST SHADOWS ALL
EXPOSE {shadowcopy-id} S:
5. Windows Server 2008 MultiPath I/O (MPIO)
5A. Goal
Making sure system uses both paths and that fail of one path can be tolerated.
5B. Setup
Network: Make sure hosts, switches and SAN are properly connected. Make sure there are two connections between the host and the Storage network.
SAN: Configure device.
Host: Load/configure HBA drivers or iSCSI Initiator. Load VDS Hardware Provider. In Server Manager, add the Storage Manager for SANs feature.
5C. Test
In Storage Manager for SANS (SMfS), create a LUN, assign it to the server and create a volume on it.
Copy a files to the volume.
Confirm that both paths are enable and are being used.
Disconnect the cable in one of the paths.
Copy a files to the volume.
Confirm that surviving path is working properly.
Reconnect the cable to re-enable the second path.
Copy a files to the volume.
Confirm that both paths are enable and are being used.
6. Windows Server 2008 Failover Clustering (WSFC) Validation
6A. Goal
Making sure systems are ready for Windows Server Failover Clustering.
6B. Setup
Network: Make sure hosts, switches and SAN are properly connected. Make sure Ethernet network has two paths.
SAN: Configure device.
Hosts (two): Load/configure HBA drivers or iSCSI Initiator on every node. Make sure all nodes are a member of the same domain. In Server Manager, add the Storage Manager for SANs feature and the Failover Clustering feature.
6C. Test
In Storage Manager for SANS (SMfS), create two LUNs, assign them to both servers and create a volume on each.
In Failover Cluster Management, from the first node, run “Validation a Configuration”.
Confirm everything is properly configured by reviewing the Validation report.
7. Windows Server 2008 Failover Clustering (WSFC) with MPIO
7A. Goal
Making sure system uses both paths and that fail of one path can be tolerated in a Failover Clustering environment.
7B. Setup
Network: Make sure hosts, switches and SAN are properly connected. Make sure Ethernet network has two paths. Make sure there are two connections between each host and the Storage network.
SAN: Configure device.
Hosts (two): Load/configure HBA drivers or iSCSI Initiator on every node, making sure MPIO is used. Make sure all nodes are a member of the same domain. In Server Manager, add the Storage Manager for SANs feature and the Failover Clustering feature.
7C. Test
In Storage Manager for SANS (SMfS), create two LUNs, assign them to both servers and create a volume on each.
In Failover Cluster Management, from the first node, run “Validation a Configuration”.
Confirm that MPIO is properly configured by reviewing the Validation report.
In Failover Cluster Management, from the first node, run “Create a Cluster”.
In Failover Cluster Management, from the first node, configure a “File Server” service.
Confirm that both paths are enable and are being used.
Disconnect the cable in one of the paths of node 1.
Confirm that surviving path is working properly.
Disconnect the cable in the second path of node 1.
Confirm that the cluster fails over to node 2.
Reconnect one of the cables to re-enable a single path in node 1.
Move “File Server” service back to node 1.
Reconnect the second cable to re-enable multiple paths in node 1.
Confirm that both paths are enable and are being used.