Gone are the days that we would use tape to bring IT systems back from a disaster. Cloud-based architecture makes the process simple and fast. Unfortunately, while complexity has decreased, cost has not. Having no plan is not an option. The question instead has become: “How do I streamline my disaster recovery plan?” These five steps can help you put an effective plan into action and balance your disaster recovery budget:
Step One: Evaluate What Needs Protecting
Start by defining what services you’re protecting. Then redefine what’s critical. Is this service necessary for business survival? Or can wait longer for recovery?
A customer from a previous Cloud DR business I ran a few years ago had a fire at their facility, and it wasn’t email or CRM that saved the company. It was a finance system. The ability to allow money to flow in and out of the organisation was the most the important factor.
Step Two: Understand the Difference Between Urgent and Important
Easy examples of urgent services are active finance, email and instant messaging applications. There is a clear need for these services across all industries in the event of a disaster. Aggressively mirroring / journaling / snapshotting these applications will give fast recovery times. Using a hypervisor allows these services to be restarted quickly—but comes at high cost.
Other services may be deemed important, but not urgent. Unstructured file data, which makes up the majority of data storage today, is one example of important but not urgent. If you protect unstructured data in the same way as other services, you achieve fast recovery times. This enables employees to concentrate on their day jobs. The storage costs, however, can be up to 3 times the initial capacity. For example, for 1TB of source data you may have, you could create an additional 3TB of copies through snapshots, mirrors and backups.
Sales teams and contracts can make things particularly tricky. A proposal or contract is absolutely critical during creation, negotiation and amendment and publishing. The documents are often created to tight deadlines. If these deadlines are missed, it could cost the business sales and revenue. That means that these services are deemed urgent. An aggressive protection policy will protect the documents at their most critical stage. Once the deal is done and the contracts are signed, these documents become reference and no longer require the same type of protection. They become important, but not urgent.
Step Three: Map the Right Technology to the Right Job
Once you understand the difference between urgent and important, you can begin to map the right technologies to the right jobs. This is important because the vast majority of the data protection / business continuity solutions are based upon capacity. If you can balance your recovery point and time objectives correctly, you can lower your costs.
- Zero downtime services demand a geographically clustered solution paired with a continuous data protection solution mirroring disk writes. Use snapshots for point- in-time protection against software disasters.
- Urgent services that only allow minutes of downtime per year need recovery points in seconds and times in minutes. Use a continuous data protection solution coupled with snapshots.
- Less urgent services allow more room for downtime. Use a snapshot to provide point-in-time rollback in case of software disaster and rely on the nightly backup to restore the full volume.
A Final Word
Tape has a place in protecting data assets. If it’s important but not urgent, you have the luxury of longer recovery times. If your data is structured correctly then you can have the luxury of choosing higher latency recovery points (weekly / monthly) using the same backup software and technology that you are using today. A proper archiving solution can further reduce the costs of disaster recovery for your company. Learn more about the connection between archiving and saving money. Download the article, here.