This section explains the installation of a Disaster Recovery environment.
Install a Disaster Recovery environment using the following flow.
Pre-setup Preparations
Perform pre-setup preparations for the primary and backup sites.
For details, refer to "Chapter 3 Resource Orchestrator Setup Design" of the "Setup Guide CE".
Replication of Storage Units
Copy the primary site information to the backup site using the storage unit replication function.
For storage replication, use the functions of "14.1.3 Storage Chassis High Availability".
Disaster Recovery can be operated on LUNs created beforehand or on dynamic LUN mirroring storage of ETERNUS. For details on prerequisites for storage replication, refer to "3.4.2 Storage Chassis High Availability Design" of the "Setup Guide CE".
For primary sites, it is not necessary for the disk resources for backup site to have been detected. Therefore, it is not necessary to connect storage units for the backup site to the server for the primary site.
Configure the image file storage folder specified during installation of Resource Orchestrator on the storage of the replication target. Copy the image files stored in the image file storage folder in the same way.
Note
When using dynamic LUN mirroring storage of ETERNUS, do not use virtual storage resources. If virtual storage resources are used, the importing performed for switchover in Disaster Recovery will fail.
When using dynamic LUN mirroring storage of ETERNUS, replacing by LUN which was made beforehand when changing to the backup site once from the primary site. It is necessary to build environment of replication of LUN made beforehand after that.
When an error occurs in the middle of transmission of online data using the replication functions of storage units, the system may not recover correctly depending on the status of data.
Also, perform backup of the primary site regularly.
For directory service configuration
When using directory services, configure the same settings for the primary and backup sites.
When creating physical L-Servers
Refer to "Appendix D Design and Configuration when Creating a Physical L-Server" of the "Setup Guide CE.
When using server virtualization products to create L-Servers
Refer to "4.4.2 Settings for Server Virtualization Software" of the "Setup Guide CE". Installation of VM management software on the backup site is not necessary because it is installed on physical L-Servers on the primary site.
When using VMware virtual L-Servers
Enter the following commands for the VM host and disable detection of the copied LUNs.
>esxcfg-advcfg -s 0 /LVM/DisallowSnapshotLUN <RETURN> |
When using the auto-configuration function for network devices
Copy the files and folders below to the backup site.
Installation_folder\SVROR\Manager\etc\customize_data\network_device_model.xml
Installation_folder\SVROR\Manager\etc\scripts\
Installation
Install Resource Orchestrator on the primary and backup sites.
For details, refer to "Chapter 5 Installation/Uninstallation" of the "Setup Guide CE".
When you change from the primary site to the backup site repeatedly, it is necessary to change the host name of a manager server every time. In addition, it is not possible to reuse a host name that you have used.
When Disaster Recovery is used, the license for the "DR option" is necessary.
After installing Resource Orchestrator, register the license for the"DR option" on the primary and backup site.
For details on how to register the license, refer to "7.1 Login" in the "Setup Guide CE".
Setup
Set up Resource Orchestrator on the primary and backup site.
For details, refer to "Chapter 8 Setup" of the "Setup Guide CE".
For Disaster Recovery environments, configure as follows.
Configure the same values for the following tuning parameters of the primary and backup sites:
Installation_folder\SVROR\Manager\rails\config\rcx\vm_guest_params.rb
For details on settings of tuning parameters, refer to "When a Virtual L-Server is Stopped, "Message number 69122" is Displayed, and Stopping of the L-Server Fails." of "4.4 General Operations for L-Servers" of the "Reference Guide (Resource Management) CE".
It is necessary to register the same resource name in the server tree and the network devices tree, and the same management software name in the management software tree for the primary and backup sites.
When you adopt automatic IP address allocation and server selection from a server pool for L-Server, it is necessary that the IP addresses that are used for servers in the primary site are assigned to eliminate overlap with the ones of servers in the backup site.
It is necessary to set the same IP address for the IP address of the admin LAN used by the admin server on the primary and backup site.
Regarding the following configurations, use the information satisfying the necessities of both the primary and backup site.
Installation_folder\SVROR\Manager\etc\storage\emc
Installation_folder\SVROR\Manager\etc\customize_data\storage_portset.rcxprop
Under Installation_folder\Manager\etc\customize_data\fc_connection_pattern
When using a virtual L-Server, it is necessary to operate the following respectively:
The server to operate VM management software on the physical L-Server
VM hosts
It is recommended to use simple names for resources to manage in order to understand the relation of resource names of the primary and backup site easily.
Certificates installed on both the primary and backup sites must be the same. When installing a manager on the backup site, copy the following files from the primary site. Perform this operation after stopping the manager.
When configuring a single backup site to inherit multiple primary sites, copy the following files of the manager on the primary site that is first created to the managers on the other primary sites.
Installation_folder\SVROR\Manager\etc\opt\FJSVssmgr\current\certificate
Installation_folder\SVROR\Manager\etc\opt\FJSVrcxdm\certificate
Installation_folder\SVROR\Manager\sys\apache\conf\ssl.crt
Installation_folder\SVROR\Manager\sys\apache\conf\ssl.key
The access rights that were modified on the primary site are not carried over to the backup site. Update them manually when the switchover has completed.
Configure the server NIC definition
When using a NIC other than one in the default configurations of network auto-configuration and using a blade server, create a server NIC definition, and reflect it on the manager.
Create the same server NIC definition for the backup site same as for the primary site in advance, and reflect it on the manager.
Execute the rcxadm nicdefctl commit command, and reflect the created server NIC definition on the manager.
For details on the server NIC definition, refer to "2.11 Server NIC Definitions" of the "Reference Guide (Resource Management) CE".
For details on the rcxadm nicdefctl commit command, refer to "1.7.16 rcxadm nicdefctl" in the "Reference Guide (Resource Management) CE".
