DataCore splits disaster recovery load among remote sites with Advanced Site Recovery

DataCore's Advanced Site Recovery allows organizations without a secondary DR site to split workloads among remote and branch offices (ROBOs), then fail back to a central location.

DataCore Software Inc. rolled out new software at the Microsoft Worldwide Partner Conference that allows organizations without a secondary data center devoted to disaster recovery (DR) to fail over physical and virtual servers to multiple remote or branch offices (ROBOs) in the event of a disaster.

The new software, called Advanced Site Recovery, or ASR, requires DataCore's SANmelody or SANsymphony storage virtualization software to be installed on all storage nodes involved in the disaster recovery plan. The customer would also install Advanced Site Recovery on each DataCore node and decide which remote sites should take over each workload in case of failure.

This requires detailed planning and knowledge of the capacity available at each remote site, but DataCore director of product marketing Augie Gonzalez pointed out that the software will only be available through DataCore-certified Advanced Site Recovery implementation channel partners, who will assist with the planning and deployment of Advanced Site Recovery.

More on disaster recovery planning and management
Disaster recovery strategies for virtual machines

Recovery time objective and recovery point objective in disaster recovery planning

Testing IT disaster recovery plans

Verizon Business aims to back up and restore data without overburdening networks

 Customers also preconfigure how to stand up new physical or virtual machines at the remote site in the event of a failure -- either by mounting a physical server to a stored system image or doing a physical to virtual server conversion. To keep data up to date at secondary locations, DataCore will replicate and store periodic snapshots, including system-state images that can be used for a bare-metal reboot of physical servers.

When a disaster is declared, customers can press a button to kick off Advanced Site Recovery. ASR will fail over the hosts, re-address the secondary sites with IP addresses from the main data center and redirect users. After the disaster or disruption, ASR can fail back to the main data center either in a staggered restoration one workload at a time or all at once, also according to the customer's pre-configuration.

Advanced Site Recovery vs. VMware

Advanced Site Recovery is similar to what VMware has been working on with distance VMotion. But one element VMware doesn't yet have covered is the replication of data on storage servers as well as virtual machines. Distance VMotion also doesn't apply to physical machines, and both distance VMotion and VMware Site Recovery Manager (SRM) require a dedicated secondary data center. Distance VMotion is also at a proof-of-concept stage, while ASR is available.

"It's a classic challenge in the VMware world," said Rick Villars, vice president, storage systems and executive strategies, IDC, of the comparison between VMware's proof of concept and DataCore's shipping product. He predicts some organizations will probably wait for VMware regardless. "There are some people who want to live in the VMware world and want everything managed through VMware," he said.

The concept of automated failover among multiple sites isn't brand new, especially in the mainframe world, pointed out 451 Group vice president of research operations Dan Kusnetzky. "What's new is this capability becoming available on industry-standard hardware with industry-standard microprocessors," he said. All of that has its roots in server virtualization, but "as we move to a virtualized environment it's critical to consider storage virtualization, as well as ways to do access virtualization, so end users don't necessarily know where the application and data store they're talking to are located."

While there is a lot of planning involved in implementing Advanced Site Recovery, "It's a different order of magnitude to set up a dedicated disaster recovery data center, hiring staff, etc.," Kusnetzky said. "And any kind of disaster recovery requires planning and ongoing testing."

Also, while remote sites can take over the organization's workloads automatically, Gonzalez acknowledged there's inherent network latency introduced when data center workloads are ported to remote sites that may be on the other side of the country or the globe.

However, he also pointed out that splitting workloads among multiple sites adds parallelization -- DR data centers don't usually have the same networking circuits as primary sites. "It's a matter of going from a 45-megabit circuit to a bonded 3-megabit circuit at a single DR data center versus going from 45-megabits to 12-megabits spread across four different sites," he said.

Still, Kusnetzky predicts Advanced Site Recovery will be most appealing to shops which currently don't have a disaster recovery plan. "As people start to think about disaster recovery, they can consider resources that might be idle, and introduce DR in a really gentle way," he said. "Larger organizations would probably consider tools from different vendors, but they also might like this for certain smaller sites."

"For a lot of midsized organizations, disaster recovery without a secondary data center is their biggest challenge," Villars said. With Advanced Site Recovery, "there are a lot of moving parts to coordinate, but the DataCore channel base is considered very sophisticated."

Advanced Site Recovery is available now from DataCore channel partners, for a list price of $2,000 per site.


Dig Deeper on Disaster recovery planning - management

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.