Milagro Exploration used Nexus 7000 Overlay Transport Virtualization feature to extend LAN across
200 kilometers.
Business Challenge
Based in Houston, Texas, Milagro Exploration, LLC is a privately held independent oil and gas company, primarily engaged in acquisition, development, and production. The company's approximately 1000 producing wells span a 600-mile stretch of Gulf Coast, where hurricanes are inevitable. Business interruptions from weather-related outages or disasters are unacceptable, however, because Milagro needs to comply with strict regulatory and reporting requirements.
In late 2007, Milagro acquired another company, increasing overnight from 15 employees to more than 200. "We're careful with expenditures, but we recognized that a major investment in disaster recovery would be critical to achieving our ambitious growth plans," says Jim Ivey, CEO for Milagro.
Milagro began looking for a disaster recovery solution that was highly scalable, could grow with predictable costs, and minimized management overhead for the three-person IT team. "Although we're not a big company, we have a big company's needs and strive to automate as much as possible," says Aaron Seward, director of IT for Milagro Exploration.
Solution and Results
Milagro is now able to replicate 20 TB of data from the Houston data center to a collocation facility 200 kilometers away in Austin, Texas, via snapshots performed approximately every 15 minutes. The disaster recovery solution is based on Cisco Nexus® 7010 Switches, the Overlay Transport Virtualization (OTV) feature, and NetApp SnapMirror software. OTV enables Milagro to extend the Ethernet LAN across data centers using a Gigabit Ethernet VLAN connection from its carrier. "The Cisco Nexus 7000 was the clear best choice in our environment because of the combination of OTV, scalability, and flexibility to support multiple protocols in one switch," says William Walla, manager of network infrastructure for Milagro.
The Houston data center and Austin collocation facility each have a Cisco® Nexus 7010 Switch configured with two separate virtual device contexts (VDCs). One VDC in each switch connects to the switch in the other data center, using OTV. The other VDC in each switch connects to the network core and virtual servers.
The major benefits of the Cisco Nexus 7010 Switch for Milagro include:
• Business continuity: Milagro anticipates achieving a one-hour Recovery Point Objective (RPO), thanks to the high performance of the Cisco Nexus 7010 Switch and NetApp SnapMirror technology. The IT team saw the impact of the switch's high performance after using NetApp SnapMirror to replicate a 10 TB volume between data centers in just 19 hours. Replicating the same volume within the Houston data center takes 30 hours, more than 50 percent longer. Another Cisco Nexus feature that contributes to business continuity is the In-Service Software Upgrade (ISSU) capability.
• Faster and simplified disaster recovery: "OTV enables us to move our server VLAN from Houston to Austin with a couple of commands, without client-side changes," says Seward. The IT team estimates that the replicated data will meet 85 to 95 percent of daily business needs.
• Predictable costs as business grows: The Cisco Nexus 7010 Switch has a lower five-year total cost of ownership for Milagro. "I was able to justify the expense to our investors, because it makes the cost of business growth more predictable," Ivey says. For example, high port density enables a single switch module to support more servers, and the 10 Gigabit Ethernet capability, which Milagro plans to use later, will support more virtual machines in each blade server enclosure.
• Consolidation, for lower total cost of ownership: The other disaster recovery architecture that Milagro considered would have required 28 stackable switches in each data center, each with its own copper cable. The Cisco Nexus 7010 solution requires just one switch with two cables, significantly simplifying management. "Data center space is very valuable, and we can use the saved space to add blade server enclosures as we continue to grow," says Seward.
• High performance and low latency: "OTV doesn't add any latency, enabling us to locate our disaster recovery data center farther away," says Cody Fleming, manager of enterprise operations. Low latency will enable Milagro to use VMware Site Recovery Manager, a future plan.
• Low training requirements: The NX-OS used in the Cisco Nexus family of switches closely resembles the Cisco IOS® Software that the Milagro IT team uses to manage Cisco Catalyst® Switches.
As Milagro continues to add more data and virtualized servers, the IT team plans to add Cisco Nexus 5000 Series Switches at the distribution layer and Cisco Nexus 2000 Series Fabric Extenders at the server access layer. Cisco Nexus 1000V Software Switches will help ensure that each virtual machine's networking and security policies travel with it as it moves to different blade servers and across data centers.
Seward concludes, "The Cisco Nexus platform was easy to implement and provides much better performance and disaster recovery capabilities than other options. In the long term, the Cisco Nexus 7010 Switch will simplify business growth, because we can connect to any fabric, and to physical as well as virtual servers."
For More Information
To find out more about the Cisco Nexus family of switches, visit: www.cisco.com/go/nexus.
To find out more about Cisco Data Center Business Advantage solutions, visit: www.cisco.com/go/dc.
