By Sungard AS
As an individual, you know you're living in a digital age when you're more concerned about crashing your computer than crashing your car. For businesses, it's no different: data sets are the crown jewels of any enterprise, and that means a huge emphasis on data recovery plans.
But data recovery is hardly straightforward for any organization. There are innumerable pitfalls that can have a devastating effect on your business if you should experience a data recovery situation.
To save you heartaches, headaches, and lost data – here are the 5 things you should know about data recovery plans.
1. Choose a credible data storage vendor
Data storage is the foundation for data recovery. If you want to partner with a data storage vendor, know that, as with any product or service, there are good vendors and questionable ones. Some data storage vendors use inexpensive, unreliable disks. They copy data, but don't replicate it. They may not make a true secondary copy, but instead use "pointers" to the original data set (which is fine, unless the original data set becomes contaminated). When disaster strikes, you suddenly find that failing over is a pipe dream – your vendor doesn't actually provide that capability.
A credible data storage vendor, on the other hand, provides an end-to-end disaster recovery (DR) ecosystem. This entails:
- Both backup and replication (either synchronous or asynchronous) to ensure that you have an up-to-date copy of all your data
- A complete secondary copy, with roll back capabilities, of all data to guard against data corruption
- Data encryption to protect data capture, data in flight, and data at rest
- Comprehensive failover capabilities to support business resiliency
Remember: you want to be able to get to your data anytime, anywhere, under any circumstances. That is what data storage is for.
2. Purchase a new storage array with your data replication software
Let's assume that you want to handle data replication and storage yourself. You would think – logically speaking – that it would cost more to buy a storage array AND replication software than to just buy replication software and re-use your current hardware. Not so.
For some unknown reason, it is typically cheaper to purchase a new storage array along with your replication software. So just do it. You will benefit from the latest storage array technologies with regard to drive density and configurations, so you're not losing anything. Additionally, the latest replication software may not run well on older hardware, and the maintenance for older hardware may become cost-prohibitive.
Let your decision on what to buy be guided by the replication software you want. That's the most crucial piece of the puzzle, because the software is what ensures your ability to recover your data. Then, buy the recommended storage array along with the software license.
3. Choose between file and block storage protocols
Data recovery cannot be divorced from your day-to-day business operations. That is why the first question I always ask when someone says "I want storage" is "File or block?" This has to do with how your data will move from Point A (your production applications) to Point B (data storage).
File-type protocols use the existing network connections, which make them very easy and inexpensive to use. All you need to do is plug in to your network and hang the storage device off of it. The drawback is that file-type protocols can chew up all your network bandwidth. So, if someone needs to send a terabyte of data to the storage, it can slow down everybody in the company.
In a block scenario, you use a different type of cable called a fiber channel. You also need a whole separate infrastructure dedicated to super fast transactions. It's expensive, but with block protocols, you don't slow down your network.
Often, startup companies choose file protocols because of the cost efficiencies, but then move to block protocols as they mature and data transactions begin to mount up.
4. Select a reliable network carrier
Choosing a network carrier is important because you want to make sure your wide area network (WAN) circuits are always on – otherwise, your data will never make it into storage and recovery won't be able to happen. Look for:
- Service level agreements that guarantee availability
- Real-time monitoring and troubleshooting
- Redundant network circuits for re-routing traffic
5. Consider business continuity and disaster recovery (BC/DR) storage input/output per second (IOPS)
Finally, decide whether you want your BC/DR storage IOPS to be as fast as your production environment (which will cost more), or whether it can be slower (which will be cheaper). You can look at it this way: you want a second car in your garage in case your primary car has a problem. But if your primary car is a Ferrari, does your backup car also have to be a sports car? Or would a Chrysler 300 be sufficient?
By taking the above factors into account, you will create the data recovery plan you need to protect your company's crown jewels!
This article was originally posted on Forbes BrandVoice.
Learn more about Disaster Recovery and Business Continuity services from Sungard Availability Services.