Data loss and the “I’m alright” fallacy

How your business’s survival could depend on resilient storage

We’ve all heard the stories: small and mid-size businesses losing their data and suffering severe financial loss, or in a few cases even going bust. The temptation is to think of them as outliers – as rare unfortunates. After all, that could never happen here, could it? We’re protected, surely – and anyway, we don’t depend on our computers and our data storage that much! But we do – we all do. We live in the CRM, on email, or in some other business-focused application that we would be lost without. Software permeates almost every area of the business, it also brings efficiencies and lets the business do more with fewer people. However, it also means that reverting to manual operations would be very difficult in some areas of the business, and perhaps impossible. As more and more processes go digital, this challenge is only going to grow. At the same time, that essential software is utterly dependent on servers and networks, and they in turn rely entirely on their underlying data storage. All of that could vanish in a moment in the event of a catastrophic storage failure – and it doesn’t even have to be catastrophic. Far more likely is data loss or corruption from a user error or software bug – or from something malicious, such as ransomware. “But we have backups,” you cry. The trouble is that restoring from backups can take considerable time, even assuming they are valid, complete and up to date. Many are the stories of companies discovering too late that the recovery process had never been tested, so no one realised that the tape was full or worn out, that some data had been left out of the loop, or that the only usable backup was an old one and more recent data had to be re-keyed. And of course, if a storage device has failed, you might have to wait for new hardware to arrive before you can restore the data. Unplanned replacements or upgrades can be costly and disruptive. It really is much better to have more robust systems to start with. So how can an SMB build a data infrastructure resilient enough to support a modern business? The first thing to remember is that there are three layered elements to resilience: High availability: this keeps systems running even after a hardware failure. Typically data is mirrored onto another disk or storage system, and if the primary copy fails the application automatically switches over to the spare. A key caveat is that this provides no data protection as such – if you delete or corrupt data on the primary, that is automatically replicated to the spare.  Disaster recovery: this is the one everybody’s familiar with – your data centre has flooded or burnt down, or your storage has totally failed or been corrupted by ransomware, and you need to rebuild your systems. It’s the main reason for doing replication and off-site backups. Data protection: a lower profile but more common risk is the loss or corruption of individual files or databases. It may be possible to recover the data from a backup, but that can be a time-consuming process, especially if you are still relying on tape backups or if your backup data is held “in the cloud” and you have to pay to get it back or if you don’t have a lot of network bandwidth. Specific data protection technologies will do a better job, for example snapshots to roll a system or service back to an earlier known-good state, cloud backup services, and backup appliances that include a file archiving element that provide users with self-service access to past versions of their data. Fortunately, the evolution of storage in the last few years has brought considerable advances. Not only are there consolidated off-the-shelf ‘data protection appliances’ that combine multiple layers of protection, but the same evolution has taken place on the high availability side too. Perhaps more importantly, these technologies are increasingly available at SMB-friendly prices. Several developments have made this possible, including the inclusion of enterprise-grade flash storage and the inclusion as standard of software technologies that were once payable extras. For example, with the right choice of modern storage, an SMB should easily be able to benefit from:

  • Replication, using ‘stretch clustering’ to transparently fail-over to a secondary system or site;
  • Snapshots, for application-consistent recovery of both virtual and physical systems;
  • Virtual volumes, integrating with technology such as VMware VVOL for better virtual machine support;
  • Multiple media support, including flash, 2.5” and 3.5” disks, and future solid-state storage such as NVMe;
  • Comprehensive data management, including automated QoS controls, even on entry-level systems;
  • Integration with data protection appliances and/or cloud backup for tasks such as file recovery.

Traditionally, putting all this together was a complex integration job. Multiple hardware and software products had to be acquired and configured, which was a complex task and meant that the only people able to do the job properly were larger organisations with skilled staff, or those with the funds to pay a system integrator or specialist services company to do it for them. That’s no longer the case. Some mainstream storage appliances now provide integrated and automated data availability and disaster resistance, offering SMBs the same top-grade ‘five nines’ protection that was once the preserve of large enterprises in a device that’s simple to implement and operate. Put it all together in the right package, and the potential for improving operational efficiency is immense.

Click here for more posts from this author

Bryan Betts is sadly no longing with us. He worked as an analyst at Freeform Dynamics between July 2016 and February 2024, when he tragically passed away following an unexpected illness. We are proud to continue to host Bryan’s work as a tribute to his great contribution to the IT industry.