Here we will discuss about 2 node failover clustering.
To do this first we need to know prerequisite or lab configuration:
· One Domain machine Test.com (IP- 10.10.10.4)
· Two member server machines those should be member of test.com with two network interface.
§ Cluster1: Cluster1.test.com
o Public IP: 10.10.10.5
o Private IP: 192.168.1.5
§ Cluster2: Cluster2.test.com
o Public IP: 10.10.10.6
o Private IP:192.168.1.6
· One iSCSI Target machine (IP:10.10.10.254)[for cluster disks]
Promote pc1 to a Domain controller:
Pc1 is a win server 2008 machine Install ADDS Role and promote this as a test.com.
Install and configure Exchange Mailbox, Hub Transport, and Client Access roles on it.
Configure Target Machine:
To configure Target machine see another post how to Configure Target machine.
Configure Cluster machines:
Configure Network properties-
Step1: configure Public Network interface properties
Step2: Click on advanced and configure DNS and WINS properties on public interface.
Step3: Configure Private Network (Heartbeat) interface Properties.
Step4: Click on advanced and Configure DNS and WINS properties on private interface.
Step5: Disable file and printer sharing on private network interface under advanced option.(this option may not applicable for ccr)
Now we have done network configuration.
Method 1: Node and disk majority quorum –
Configure witness disk for quorum (Single Copy Cluster)
Step6: Add SAN device to cluster storage on each node which will become Failover clustere. To do this Configure iSCSI Initiator Properties= Go to Start>Administrative Tool>iSCSI Initiator> Click on Discovery tab>add portal>add Target ISCSI server IP.
Step7: Click on Targets Tab> click to log on button
Step8: Select Automatically Restore this connection when the computer starts> click ok.
Step9: Realize that all targets are connected, If not connect to all one by one.
Step10: Click on Volume and Devices tab> click on Autoconfigure and click ok.
Step11:Initialise disk and choose partition style MBR.If disk Size will larger than 2TB, it will must to Choose Partition style GPT.
Step12: Create new simple volume
Step13: choose NTFS file system, it is recommended.
Step14: Finish creating new volume.
Tep15: Repeat step12 to step14 to create partition on each Disk.
Step16: Shutdown cluster1 machine or disable all network interface before configuring second cluster machine to prevent data corruption.
Repeat step1 to step10 on cluster2 machine and configure with different IP, after that go to Server Manager> Storage> Disk manager> select Disks on by one and initialize it.
Create Failover Cluster:
Step16: open CMD and type a command line parameter to install failover clustering feature on both node:
ServerManagerCmd -i Failover-Clustering
Step17: Go to Start>Administrative Tool>Failover Cluster Management>Create a Cluster.
Step18: Add Cluster node addresses.
Step19: Add a cluster Name (Users will use this name to connect Services and Application on the server) and a New IP.
Step20: Face of create cluster summary page.
Step21: summary page of new cluster.
Step22: You can see below cluster node is online.
Step23: Look all clustered disks are online or not?
Step24: Go to networks and see Allow clients selected or not under public network properties.
Step25: To validate configuration click on validate configuration.
Step26: Add cluster name> click Next.
Step27: Select Run all test> click Next.
Step28: On confirmation page click next.
Step29: Here Validation process is running.
Step30: Click on report to see overall validation Report.
Step31: To Realize Cluster working or not, Go to cluster 1 which is currently online and shutdown it. After that go to Failover Cluster management console you will see an error on summary page.
Step32: Also you can see Unavailable Status on Network interface.
Step33: You can see here cluster2 is become quorum owner automatically mean that cluster is working fine.
Method2: Node and File Share majority Quorum-
To create file share witness for quorum [Cluster continuous replication (CCR)]
Note that CCR configuration not need SAN devices. If u install cluster without SCSI device that’s quorum configuration will set to node majority by default.
Step34: create the file share on a separate server (If u paln a CMS for echange server, best practice is to use a Hub Transport server in the same AD site as the failover cluster nodes. To do so, log on to the server in your lab environment on which the Client Access and Hub Transport server roles are installed.)
To make a new directory run cmd command:
Note: Microsoft recommends using the following naming convention for the directory name: FSM_DIR_
To create the share by running the following command:
E:\>net share fsm_dir_ccrmbx=e:\fsm_dir_ccrmbx /grant:$,full
Note: Microsoft recommends using the following naming convention for the share name: FSM_
Note that only the Windows 2008 failover cluster account gets permissions to access this share. Also notice that the failover cluster machine account is hidden, meaning you must add a “$” sign after the name.
Now Assign file system permissions to the directory being shared by running the following command:
E:\>cacls e:\fsm_dir_ccrmbx /g builtin\administrators:f $:f
Note: If u got share folder permission error permit full access to both node .
Either you can install cluster using CMD:
• Initiate Cluster service
cluster WINCLUSTER /create /ipaddress:10.0.0.101 /User:domain\ServiceAccount/password:<> /verbose /unattended /minimum
• Make the cluster an MNS cluster
cluster /cluster:WINCLUSTER resource MNS /create /group:”Cluster Group”/type:”Majority Node Set” /privMNSFileShare=”\\mail\fsm_dir_mbx”/online
• Delete the default Local Quorum
cluster /cluster:WINCLUSTER resource “Local Quorum” /Offline
cluster /cluster:WINCLUSTER resource “Local Quorum” /delete
• Add nodes to the cluster
ocluster /cluster:WINCLUSTER /Addnodes:NODE1 /Minimum/Password:<>/verbose
ocluster /cluster:WINCLUSTER /Addnodes:NODE2 /Minimum/Password:<>/verbose
Or install using GUI:
Step35: With the system file share created, we must configure the quorum settings for the failover cluster, so that they point to the share we just created. To do so, open the Failover Cluster Manager on one of the nodes> cluster name> Action>More Actions> select “Configure Cluster Quorum Settings” in the context menu.
Step36: Add quorum shared folder UNC path.
Step37: After configuring cluster NFSM u can view summary will look like this.
Step38: Note one thing here none of these cluster disks are used for quorum.
Step39: In the Failover Cluster Management tool> click the clustered mailbox server> in the details pane, click Cluster Core Resources> Double-click the File Share Witness resource> Click the Policies tab.
In the If all the restart attempts fail, being restarted again after the specified period (hh:mm) box, type 00:15.
This action changes the cluster file share witness restart time from the default restart time of one hour to 15 minutes.
Click OK to apply the changes, and then exit the Failover Cluster Management tool.
Step40: validate cluster configuration
Step41: Now we can install different services or application on it to be clustered.