5 Steps for Creating Cost-Optimized Disaster Recovery Plans

By Chris Poelker

Companies cannot afford to have any downtime in todays 24/7 economy. When channel partners ask their customers if they have the technologies to properly back up and recover data almost instantly, the answer is not always a resounding, Yes.” In fact, many CIOs and IT managers say they lie awake at night asking themselves what would happen to the business if there were a disaster. What would the company do if one of the application servers died, or, even worse, the entire data center lost power? How would that affect the services the company provides to clients and, ultimately, how would the business be impacted? Is there a plan to ensure business continuity if a natural or man-made disaster occurs? Companies often don’t have good answers to these questions and that puts their businesses at risk. But, they are turning to their channel partners and managed service providers (MSPs) for help.

MSPs and channel partners shed the light on why backup and disaster recovery (DR) plans are like insurance policies. DR plans dont enhance a companys mission or its bottom line, but without them, the companys mission and bottom line are at stake. Plans may be annoying, but they are necessary. For example, MSPs’ and channel partners’ customers seem to think that if they have IT professionals managing the enterprise data centers and infrastructure, then they have a DR plan. But, deep down, they know that if their company actually flooded, they would be in serious trouble. That’s also the case when the traditional DR plan is to ship tape off-site. When that process is used, the time, expense and complexity of actual recovery is understood to be weeks or months. If a disaster occurred, the affected company most likely would have to shut its doors. Thus, end users are turning to their channel partners for help in creating effective and efficient processes. Business managers know they can no longer afford not to care or operate in denial. Therefore, here is a step-by-step guide that resellers, integrators and other channel partners can use to help customers build optimized models for business continuity.

Step 1: Suggest adding data deduplication to reduce expenses for data storage and replication, and, in many cases, make it cheaper than shipping tape.

To prove this, assume a company is working with standard backup software without dedupe, LTO3 drives and 20 terabytes of production data. Without even considering the cost of the backup software, the price to provide 40TB of tape capacity comes out to about $66,480 (drives, media, off-site storage, recall costs, operations, etc.).

Now, explain that it’s more useful to use a single 4TB dedupe appliance that should hold the same 40TB at a 10:1 ratio. If the cost of the appliance with its lower capital and operational cost is $17,700, that’s a savings of more than 300 percent of the cost of the tape solution ($66,480 less $17,700 is $48,780, or 300+ percent in savings).

Then, suggest another deduplication appliance to replicate the data off-site. This will eliminate off-site tape movement, recall costs, any array-based replication licenses for the data, and reduce the WAN requirement to replicate that data by 90 percent with only a 10:1 dedupe ratio. That the DR side now has the data available on disk for recovery can speed the time to find, mount and recover the data, simplifying and reducing the total time from days or weeks to hours.

Step 2: Propose implementing snapshot technology to eliminate the bulk data movement for backup and improve the recovery point objectives for both backup and DR.

Backup is a copy. Physical data must be moved from Point A to Point B to copy the data. When using snapshot-based backup, backup occurs more often as the data do not need to be moved. The physics of the process is eliminated. This means data recovery is typically much faster. The backup window can be reduced to just a couple seconds as the snapshots take place.

Step 3: Offer continuous data protection (CDP) to speed recovery and eliminate data loss. Assessing how much data each business can afford to lose will determine whether this step is right for each company.

If snapshots improve the recovery point objectives from the typical 24 hours over conventional backups, CDP eliminates backup all together. It further improves the recovery point to zero data loss. Companies can afford to lose some data for some applications and for these, data snapshots are enough. But for applications running a hospital, nuclear power plant or air traffic control, for example, losing data may mean losing lives.

Even for a company that writes software, an outage may lose hours of irreplaceable coding. For situations such as these, CDP offers the best recovery service level in the industry. Data are backed up while being created. In addition, CDP can diminish the time to recover to seconds.

Step 4: Use WAN optimization and delta versioning with encryption to reduce risk and WAN requirements by 90 percent for data that have not been deduplicated.

The level of data granularity used by the replication process can have a dramatic effect on bandwidth requirements for data replication. WAN optimization reduces the granularity level from a typical 64KB or more for each write operation to individual blocks or even sectors of data per change. Some replication solutions ship a lot of white space to the other side. Delta versioning cuts down that white space to the byte level, which reduces the amount of data needing to be transferred over the WAN by 80 percent or more.

Step 5: Advise virtualizing storage and servers. This step will reduce DR infrastructure and its associated costs.

There are a number of reasons to virtualize:

Every day, channel partners encounter IT managers who are asking themselves, Am I getting the most from my current DR plan? Are my data successfully protected?” If the answer is No,” then this guide may help the channel work with customers to build a DR plan that will ensure the resiliency of critical business applications which just may be the foundation of a successful business relationship.

Christopher Poelker is vice president of enterprise solutions for

FalconStor

, a data protection and virtualization provider. He is a highly regarded storage expert with decades of experience at companies including DEC, HDS and Compaq. Poelker recently served as deputy commission of the TechAmerica Foundation Commission on the Leadership Opportunity in U.S. Deployment of the Cloud (CLOUD²).