"With PlateSpin Forge, we don't need to duplicate our production hardware environment or worry about having identical SANs in each data centre: everything we need to protect our business-critical servers is in a single, compact physical frame. The solution takes away all the complexity and unknown factors, and enables us to prove to the various business unit managers that we can get them back online very rapidly in the event of a disaster. That's a major boost for the business, and for our reputation as a service provider."
Senior Systems Engineer
With a market capitalisation of R6 billion, Metropolitan is the fourth largest listed life insurer in South Africa. The company has 72 offices employing approximately 9,800 people throughout South Africa, Namibia, Botswana, Kenya, Ghana, Nigeria, Swaziland and Lesotho. Targeting primarily the lower and middle-income markets, Metropolitan insures the lives of some 4.3 million South Africans, and provides affordable financial solutions for many more, with the goal of building prosperity for the people of Africa.
Metropolitan wanted to improve the speed and reliability of recovery from server failure, and to enhance its ability to run disaster recovery tests. The company compiled a list of "day one" servers: defined as those for which the business could tolerate no more than 24 hours downtime. These servers, running Microsoft Windows, included five Microsoft Active Directory environments and a number of Microsoft SQL Server instances.
The Active Directory servers are business-critical in the sense that they are required for the normal operation of the network and user security settings, while the SQL Server databases provide crucial support to Metropolitan's financial asset management teams.
Metropolitan's existing disaster recovery processes depended significantly on manual input from IT specialists, and the results were difficult to predict. Some Active Directory servers would come back up in 20 minutes, while others might require up to two days of reconfiguration work.
Aiming to accelerate recovery and reduce the need for manual intervention during a disaster scenario, Metropolitan reviewed analyst reports and consulted with its IT partners before shortlisting three options for protecting its physical and virtual server workload. Following an exhaustive proof-of-concept exercise, in which each solution was scored according to 15 weighted selection criteria, Metropolitan selected PlateSpin Forge®. PlateSpin Forge is a hardware appliance that provides out-of-the-box protection for up to 25 physical or virtual server workloads.
"PlateSpin Forge was the clear winner, with particularly high marks on performance, configurability and security," said Julian Brandt, Senior Systems Engineer at Metropolitan. "We are typically a relatively conservative company when it comes to technology, but the quality and robustness of the PlateSpin solution gave us confidence to adopt this leading-edge option. The solution is easy to set up and manage through the Web-based console, and it has clearly been well tested."
Metropolitan is protecting 15 servers on the PlateSpin Forge appliance, and has recently purchased a second one for less critical servers that have proved problematic to recover in the past. "There are a number of servers where application and driver issues require excessive amounts of fixing in a DR scenario," said Brandt. "Putting these servers on PlateSpin Forge will save significant work and effort, freeing up technical staff."
An important factor in Metropolitan's selection of PlateSpin Forge was the ease of failing back over to the original hardware (or to new hardware in the event of a complete disaster). "We needed a solution that could handle not only the initial recovery, but also the return to normal production after any disaster," said Brandt. "That can be really messy and difficult with some DR solutions—with PlateSpin Forge, it's simple and completely reliable. Another important benefit is the control and visibility it gives us over the failover process."
The introduction of PlateSpin Forge at Metropolitan has enabled the company to recover business-critical services faster, more reliably, and with less administrative effort. In the past, some Active Directory servers would be back up in 20 minutes, while others could require two days of intensive effort to recover. PlateSpin Forge brings speed and predictability to the recovery process, enabling Metropolitan to get these servers running again within 40 minutes.
"We recover the servers on the PlateSpin Forge appliance, then allow our Active Directory team two hours to certify that the recovery has worked as expected," said Brandt. "Within just three hours, the corporate network is back up and running."
PlateSpin Forge provides a single Web-based interface to set up and manage recovery jobs, view history, and run simulated failovers for test purposes. This saves significant time and effort, and enables the IT function to demonstrate to the business that its services are protected and rapidly recoverable.
"With PlateSpin Forge, we don't need to duplicate our production hardware environment or worry about having identical SANs in each data centre: everything we need to protect our business-critical servers is in a single, compact physical frame," said Brandt. "The solution takes away all the complexity and unknown factors, and enables us to prove to the various business unit managers that we can get them back online very rapidly in the event of a disaster. That's a major boost for the business, and for our reputation as a service provider."