Modern Infrastructure

Understanding infrastructure and platform as a service


News Stay informed about the latest enterprise technology news and product updates.

Disaster recovery case study: Preventing runaway costs

In this disaster recovery case study, one company details how it kept costs low while building a dependable backup system for its data centers.

If you run a large data center, costs for backing up data can easily spiral out of control. Here's a disaster recovery case study to show how one company reined in the spending.

IT professional: Mahmoud Harb, site operations manager

Company:, an online job recruitment site

Problem: How to back up massive amounts of job-seeker data from data centers in Atlanta, Amsterdam and Hong Kong to a backup data center in Chicago. The goal was to replicate data quickly with a low error rate and without incurring substantial costs for a major wide area network upgrade.

"The ability to push data from our primary data center to a disaster recovery [DR] site was becoming a challenge," Harb said. "We had more users as our database grew and the amount of data that we have to keep outgrew our setup. We needed a technology that could transfer data from a primary data center to the DR data center at a reasonable cost. Data freshness for backups also went from about 30 days to less than 30 minutes."

Technology solution: Hardware- and software-based WAN optimization using Silver Peak 9700 and VRX-8 virtual appliances.

"We had tried Citrix Systems' NetScaler and F5 Networks' WANJet," Harb said.

 How WAN optimization helps disaster recovery strategy

  • Speed of data transfer. "Data has to be copied to all four data centers. And the data hops have to happen within a reasonable amount of time."
  • Cost. "We do not want to rearrange our network design or topology to fit that [new] technology. We could take advantage of our existing WAN setup and transfer the data over our existing Internet connectivity."
  • Ease of administration. "There is no need to get a top engineer to run the appliance. It is self-healing and can administer itself. With a little bit of knowledge, you can get it running."
  • Support. "We have a small crew that runs a large operation (with a ratio of about 300 servers per administrator). We don't have people that can sit on the phone and talk to vendors all day long. So support was one of the key issues."

In the end, avoided a major WAN upgrade, saving the company hundreds of thousands of dollars in disaster recovery costs in just a few months.

Article 2 of 13

Dig Deeper on Enterprise data storage strategies

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.

Get More Modern Infrastructure

Access to all of our back issues View All