A disaster recovery site build can be a huge undertaking, with multiple points of financial risk that could possibly undermine the entire project. Yet, the majority of storage pros are still interested in taking on such a project.
Because there is much more risk associated with not doing it. We’ve recently helped several customers build out disaster recovery solutions in ways that were efficient and low-cost.
Here’s how we did it. . .
1. Repurposed Equipment
Saving valuable datacenter space in your production location could fund 100% of your DR project. Consider moving, upgrading and repurposing development, testing or aging production equipment from your primary site to DR. This way you can bring in newer technology to the primary location and transition aging IT assets to the DR location. By doing so you not only reduce the IT capital expense of your DR project; you also free up data center floor space in your production location and provide an opportunity to refresh equipment at your primary location.
2. Data Storage Hardware
Data storage hardware is one of the fastest growing and most expensive pieces of the DR puzzle. There are two effective ways to reduce costs: One, consider reconfiguring and repurposing an existing IBM, EMC or NetApp array from a production location to DR. Two, consider buying a used SAN storage array that has been reconfigured to meet your requirements. By doing so you could save from $30,000 to $300,000, or more, depending on the configuration. Examples of these SAN storage configurations are listed here.
3. Connectivity – SAN & Bandwidth
For SAN connectivity, you may need FCIP routers, SAN directors, switches, HBAs and additional networking equipment at your new location, coupled with Bluecoat, Riverbed or Cisco WAN acceleration to reduce bandwidth costs. By buying and deploying used CiscoMDS or used Brocade SAN directors in the site build you can save $50,000 to $150,000, leaving you room to buy the desired WAN acceleration products and bringing you long-term, recurring savings on bandwidth.
4. Transition and Logistics
Moving equipment from production to DR can cause multiple logistical nightmares. As an insurance policy against an unforeseen outage, consider using Swing or Rental IT hardware, particularly EMC or IBM storage arrays, to migrate data and offset your move. One of our customers leveraged this technique and rented an EMC storage array for two months. He shipped the array to his production location, migrated the data, then used the rental array in production, while he shipped his primary system to its new location. This gave his team the peace of mind and timeframe to move systems at their leisure to DR, and avoid the kind of “rush job” that typically ends in failure.
5. Replication Techniques
There are many ways that you can move and protect data between datacenters, including IBM, SVC, EMC Mirrorview, Commvault, Inmage, Veeam, database clustering, log shipping and others. When designing your recovery strategy, investigate multiple options and make sure to choose the one that best meets your RPO and RTO objectives. Pricing and functionality vary widely with replication technology, so do your homework; your options may be more cost-effective than you think. Further, by coupling the replication technology with used SAN storage arrays or upgrades at a 30-70% cost reduction, you can replicate and keep more data on disk, rather than tape.
You will no doubt encounter plenty of challenges and choices as you build your Disaster Recovery facility, but hopefully these ideas will get you off to a good start.
Have you had a positive experience building out a DR site? We’d like to hear it. Contact us or comment below.