Availability and the Always-on Enterprise

By December 15, 2016Backup, Data Centre

Why Backup is Dead

Backups certainly did fit the bill at one time, but data centre needs have evolved and expanded.

Every business has experienced a significant increase in customer and partner expectations around application performance and business continuity. Response times once measured in terms of seconds are now milliseconds. Downtime measured in hours or days is now expected to be minutes (if even that).

To keep pace, many organizations are implementing modern data centres. They’re deploying new technology to drive performance such as virtualized servers and networks; modern storage; and cloud-based services such as infrastructure as a service (IaaS), recovery as a service (RaaS) and disaster recovery as a service (DRaaS). In fact, 97 percent of organizations have or will have a modern data centre within 2 years to address the 46 percent of their workloads they considered to be critical. Implementing these new technologies—as well as the adoption of cloud services—demonstrates a shift in IT thinking. Instead of focusing on just servers and data and the technology needed to back them up, IT is now considering the concerns of the

Implementing these new technologies—as well as the adoption of cloud services—demonstrates a shift in IT thinking. Instead of focusing on just servers and data and the technology needed to back them up, IT is now considering the concerns of the organisation as a whole, including service delivery, operational efficiency, reputation, and most importantly availability. The need for availability is a key driver for the modern data centre in 68 percent of

The need for availability is a key driver for the modern data centre in 68 percent of organisations. Most organizations—65 percent—need real-time interaction with critical applications and systems to meet the present demand. Essentially, they need to never be down. Until the last few years, the concept of availability was only a dream. Backups were all that was available. Making a backup copy of some data set and storing it on tape or disk was enough to feel protected. Data sets and applications were much less complex, and recovery times in terms of hours were actually acceptable. These days, with the expectations around applications and service availability and the pressure for organizations to be always-on, the stakes are higher. Organizations can no longer simply rely on age-old protection technologies and methods.

Until the last few years, the concept of availability was only a dream. Backups were all that was available. Making a backup copy of some data set and storing it on tape or disk was enough to feel protected. Data sets and applications were much less complex, and recovery times in terms of hours were actually acceptable. These days, with the expectations around applications and service availability and the pressure for organisations to be always-on, the stakes are higher. Organisations can no longer simply rely on age-old protection technologies and methods.

These days, with the expectations around applications and service availability and the pressure for organisations to be always-on, the stakes are higher. Organisations can no longer simply rely on age-old protection technologies and methods.

So, is it high time the always-on enterprise killed backup? To answer that question, look at where backups fall down. Define the gap that exists between availability needs and backup capabilities. And look at how to meet the need for availability within your organisation.

You Can’t Rely on Backups Alone

Backups were the answer when making a copy of a particular set of files or even a complete single server was considered adequate protection. Today’s modern data centre is far more complex. There are multi-tiered applications, virtual servers being moved from one host to another and the cloud handling both storage and compute functions.

With organizations considering availability, it’s no longer about simply needing to restore a single file or folder. It’s about complex processes like recovering a multi-tiered application that spans multiple servers and bringing each server into a consistent state with the others. It’s a far more involved than the restore jobs of yesteryear.

In some ways, basic backup no longer has a place in the modern data centre. As new technologies have come into use—the foremost being virtualization—you can now easily move workloads from one location to another. You can even performing maintenance during the day. The options around availability are much greater and more flexible than what backups alone provide. Even so, the idea of a backup—that is, having copies of your data—is still viable.

Even so, the idea of a backup—that is, having copies of your data—is still viable. Now data centres have moved to advanced concepts like replicating data at the disk level or entire virtual machines, both from one store to another or even one site to another. This provides both increased protection and faster recoverability.

Organisations today aren’t just looking at availability on a per-application or per-server basis. The goal is to make everything available in the event of an outage. So it’s not “we have our order processing back online, but e-mail is still down.” Now it’s essential to have the entire business back up and running, not just a few services.

Should you have an availability event, can you benefit from backup alone? Backups certainly still have a place. For example, if you’re replicating changes to a VM and the source VM is somehow corrupted, that corruption will simply get replicated. So having a backup of the critical data on that server can play a role in ensuring recoverability. However, backup as the only method is no longer an option for businesses focused on being available.

As newer technologies have emerged, the frequency of backups has also shortened. In previous years, your backup window simply couldn’t be anything less than nightly. These days, backups occur much more frequently—even during production hours. And with technology like instant VM recovery in place, the concept of restoring a backup job is somewhat obsolete.

You’ve modernised your data centre, but have you also modernised your data protection strategy? Perhaps, it’s time to do so. The process begins with understanding just how far away you are from availability before you can get there.

To read Veeam’s Whitepaper in full click here.