This article explains how to create Windows Protection Group and protect a Windows server Automatically as well as Manually.
To protect a Windows server using SafeHaven, we need to install a Local Replication Agent (LRA) on the production server. In order to complete the LRA installation, a Reboot of the production server is required. Please schedule a downtime window for LRA installation.
For Ansible LRA installation to work, make sure TCP 445 port Inbound is open on the production server by running the following command via Windows Commandline (run as administrator):
New-NetFirewallRule -DisplayName "psexec tcp 445" -Name "psexec tcp 445" -Profile Any -LocalPort 445 -Protocol TCP
This article assumes that:
- A SafeHaven cluster has been created successfully
- Both production and recovery SRNs have been registered and peered
- Proper storage has been attached and claimed on the production SRN
Create a Recovery Server Template
A Recovery Server template can make Protection group deployment easier if there are multiple production servers with same NIC and PCI slot. Please click here for a KB article to create a recovery server template.
Pre-deploy a Recovery Instance(Optional)
It is possible to pre-deploy the recovery instance to speed up the Protection Group Deployment process.
- Login to Centurylink Cloud Portal https://control.ctl.io/
- Click on the DR Datacenter.
- Click on a group where the recovery instance should be deployed.
- Click on create, then select server.
- Deploy a VM using an Ubuntu-14 template that has same NIC and PCI slot number as the production server.
Please make sure that:
a. The VM has same CPU and RAM as the source VM.
b. The recovery vlan is selected to deploy the instance
Create a Windows Protection Group
Right click on the Production SRN and select Create Protection Group.
NOTE: Always start the Protection Group creation from the Production SRN.
Select the Recovery SRN, and click Start Wizard.
Click Add a Server
Enter the Server Name, OS Family and OS Version.
Select LRA under Mirroring type.
Enter IP address.
Enter the CPU and Memory so that it matches the production server
Click Add. Click Next
Click Add again if more than one VMs are needed in the protection group.
Select Automatically Deploy and Configure New recovery Proxy Servers.
Select Use pre-deployed proxy recovery server , if the recovery server has already been deployed
If Automatic deployment of the new recovery proxy instance is selected:
a. Select a Parent Hardware Group.
b. Enter a Stub Group Name if required.
c. Select Linux as the OS base for all the Stub machines
d. Enter an Alias(maximum 6 characters) for the Recovery Server.
e. Select a Proxy Template
Note : Make sure that the Template you select has the same NIC and PCI slot number as the production/source VM.
f. Select the Installed OS on Production Server.
g. In front of "Connect all the server to", select the DR vlan. The Recovery instance will be deployed in this particular vlan.
h. Enter the administrator(or root) password that you want to be set on the Recovery server.
If pre-deployed proxy recovery server is selected:
a. Select the Recovery server instance from the Recovery Datacenter.
b. Under Mapping, click on select and select the right recovery instance after clicking on down arrow.
Enter a Protection Group Name.
Click on the VM name. Click Add Disk on the right to the VM.
Expand the VM to see the disks attached to it.
Select a disk and click on Resize Disk to change the size of the disk.
Make sure that the disk layout of the VM matches the production VM.
SafeHaven supports two types of protection types. Select a Protection type:
Local Replica : 125% of total production VM's storage on the local SRN and 125% on the DR SRN.
Local Cache : 10% of total production VM's storage on the local SRN and 125% on the DR SRN.
Please click here for more information on difference between local cache and local replica
If you selected Local Replica:
Under Production Data Center, Select a primary storage pool and checkpoint storage pool. It is completly fine to use the same pool for both primary storage and checkpoint storage.
Under Recovery Data Center, select Primary storage pool and checkpoints storage pool.
If you selected Local Cache.
Under Production Data Center, Select a local cache storage pool
Under Recovery Data Center, select Primary storage pool and checkpoints storage pool. It is completly fine to use the same pool for both primary storage and checkpoint storage.
Check Automatically Install Windows LRA if you want the SafeHaven Agent to be automatically installed on the production/source VM.
Note: Automatic Installation requires a reboot of the production server. It is advised to use this option only during maintenance window.
Leave the box unchecked if automatic installation is not required.
Note: This will Create a Protection group, Deploy a recovery instance and then configure it to boot using iSCSI target from the DR-SRN disk when ever Test-Failover/Failover is performed.
Note: It may take some time for the PG to be deployed.
If the Windows LRA agent has not been installed. Please make sure to install it on the production VM during a maintenance windows.
Link to donwload agent: https://download.safehaven.ctl.io/SH-5.0.1/safehaven_windows_driver-5.0.1.exe
Note: Agent Installation triggers a reboot of the production VM.
If pre-deployed proxy recovery server was selected while creating the PG, and the server has not been configured for iscsi boot, then:
a. Right click the Protection group that was just created.
b. Go to Intallation > Install SafeHaven makestub.
c. Verify the Protectiongroup name , and click Next.
d. Enter the root password of the VM.
e. Check all the boxes.
f. Click Next.
This will configure the recovery server to boot using iSCSI off the disks of the DR-SRN when a test-failover/failover operation is initiated.
Manually Start Replication
Note: This step is only required if Windows LRA Agent was not insalled automatially during the creation of the Protection group
Log into the Windows production VM.
Verify that the SafeHaven Agent has already been installed. If not then please install the SafeHaven Agent. Requires reboot
Download link for the Agent: https://download.safehaven.ctl.io/SH-5.0.1/safehaven_windows_driver-5.0.1.exe
After reboot, click on Start, find and launch the SafeHaven Manager.exe.
Select the default options on the first screen of the wizard (all boxes checked) and click Next.
Verify the NIC type and PCI slot number. Click Next
Enter the Local iSCSI IP address of Production SRN in front of Target Portal IP Address and click Discover. Select the appropriate ISCSI target to connect from the ISCSI target list.
Click Connect, then select Next.
A new window appears which provides the mapping from source(VMDK on production server) to destination disk(ISCSI on the SRN) along with the recommended Sync Rate, accept the default option and click Next.
NOTE: Please make sure each production virtual disk you want to protect is mapped to a corresponding iSCSI target destination disk. The size of the source and destination disk must match.
Enter the IP configuration of the recovery server. Do not use DHCP. Click Next.
Click Finish to end the wizard.
The initial replication of the production windows VM will start now. Close the command prompt window if required.