The world of IT has changed dramatically over the past few decades. At one time, employees could tolerate IT outages and slow response times – in fact, they were expected. Today, IT is central to the operations of almost every organization. Outages are not tolerated or accepted; I recently found myself getting annoyed when my connection dropped on a two-hour flight.
As people became increasingly dependent on applications, the IT industry developed new methods for protecting information and reducing recovery times. The methods, however, vary almost as much as the data itself, and I have yet to see a single technology that can satisfy all needs. Many solutions get implemented, becoming additive and more complex. This can result in a mismatch of technologies and solutions being used to protect data. Each solution is implemented to solve a particular challenge or protect an application, but it can be difficult to see where there is overlap, or if an application is exposed to data loss.
Does your arsenal of data protection solutions solve all of your needs? Do you need more options, or is your system overly complex? Let’s take a look as a few common methods of data protection, see how they vary, and consider how you can make better decisions about securing and backing up your data:
The most common form of data protection, and the one people generally think of first, is backup. Even backup systems vary, depending on the vendor and methods chosen. They can provide anything from self-service search-and-restore tools to the ability to recover an entire application along with its data at a defined point in time.
Enterprise storage vendors build multiple levels of data protection into their systems. They include snapshots or storage area network (SAN)-based backups for instant point-in-time recovery, synchronous replication for continuous data protection to a remote location, or asynchronous replication for point-in-time recovery at another location. Other software vendors provide host or even hypervisor-based continuous data protection (CBT). We’re now moving from resolving simple data loss or corruption issues to quickly recovering from a complete site outage or natural disaster. Recovery time, the levels of automation, the extent of the recovery and your ability to test all need to be part of your disaster recovery plans.
To avoid outages altogether, vendors tend to use or engineer fault-tolerant hardware. Highly available systems often use the term “five nines availability,” meaning the system will be available 99.999 percent of the time, which is less than 5.26 minutes offline per year. Some servers and enterprise applications can be configured as fully redundant clusters, which means if a hardware failure occurs, the other server will pick up the workload without (or with little) pause. This is good if you need to maximize uptime, but does little if your entire site goes offline.
The permeations of data protection solutions continue to evolve. You can find studies that show percentages organizations using disk-to-tape, disk-to-disk, disk-to-disk-to-tape, disk-to-cloud, disk-to-disk-to-cloud, and so on. Some vendors claim their methods can eliminate redundant technologies. Is this true? My opinion: sometimes. It really depends on what’s needed, your tolerance for data loss and your tolerance for recovery time. You will also see these described as recovery point objectives (RPO) and recovery time objectives (RTO).
What systems will work for you? What are you trying to protect against? Take a look at the right balance of RPO and RTO that works to support your business. When evaluating your data protection environments, you may find redundancies, and you may find gaps. Take a step back and ask the right questions: How do I recover from data corruption or loss? How do I roll back to a previous point in time? How do I recover from a site failure? What compliance requirements do I have? How long should each of the recovery processes take?
Implementing more modern methods can lead to lower costs by consolidating systems. Instead of requiring a specific technology, consider your tolerances, cost of downtime, and other related requirements. You may find easier, more straightforward methods than your current implementation.
Learn more about making the most of your storage resources.