With the introduction last week of XenServer 5.6, an exciting new feature was added that addresses one of the top reasons customers are moving to server virtualization solutions – Customers wanting improved business continuity can now use Citrix StorageLink Site Recovery for XenServer Virtual Machines. This new functionality is part of StorageLink v2.2 included with XenServer 5.6. This a disaster recovery feature which allows customers to automatically create warm standby sites for their VMs. Because we have extended StorageLink to leverage storage vendors replication services to provide this feature, you will benefit by utilizing your existing investment in storage infrastructure.
Site Recovery’s wizard-based workflows dramatically simplifies the setup, configuration, and automation plans for disaster recovery for workloads running in your VMs on XenServer, providing a single user experience across heterogeneous storage solutions supported by StorageLink. Customers create Recovery Plans that allow you to group Virtual Machines that deliver an application workload together. Within a Recovery Plan, you can set up dependencies between the Virtual Machines, define to which subset of hosts on your Secondary Site the Recovery Plan should failover to, and you can specify which networks should be used on the Secondary Site for production traffic. Site Recovery provides single click access to initiate failover of your server workloads, as well as a comprehensive mechanism to test your recovery plan to validate stated recovery point objectives (RPO) and recovery time objectives (RTO) objectives.
Under the covers, StorageLink is communicating with the various storage arrays through what we call StorageLink Storage Adapters. The adapter contains specific code to manage a particular array, so they provide very tight integration with your existing storage infrastructure. For Site Recovery specifically, StorageLink makes use of replication techniques provided by the storage arrays, which once again are managed through the Storage Adapter. In the case of for instance the HP P4000 family of arrays, StorageLink will utilize the Remote Copy functionality in the array by automatically creating Remote Snapshots at the moment you start protecting your Virtual Machine data. When the customer issues a failover, StorageLink will promote the Remote Volume into a Primary Volume and utilize that to reconstruct the Virtual Machine with. Similarly for Netapp – the Netapp adapter is using SnapMirror and FlexClone technology to provide remote replication functionality.
This means that for this version of Site Recovery, you need similar arrays on either site. We are currently investigating the integration with replication services that run outside of the storage arrays, but it’s too early to go into specifics on that.
We are planning to release some new videos over the next few weeks that will give you a demo of setting up Site Recovery and how easy it is to failover workloads to your stand-by datacenter. The videos that are currently posted are made with our v2.0 release, and we have made many workflow and usability improvements since then. I will keep you posted!