As recent months have amply shown, Australia is no stranger to natural disasters. And when catastrophies strike, IT decision makers must ensure they are able to recover data quickly to get things moving again. JULIA TALEVSKI reports. When In the course of history, few have ever described the Australian climate as easy. Natural disasters such as bushfires, floods and cyclones, as horrid as they can be, are common and unpredictable. Disaster can strike anytime and anywhere. And for businesses, particularly the IT department, one of the most concerning factors is around recovering vital data and ensuring business continuity as quickly as possible. With data growing at a petabyte rate, enforcing a successful disaster recovery (DR) plan is critical when things go bad. When you consider how valuable data is and how much it will cost a business if it can’t function properly, or sometimes not all, if the data recovery mechanism takes more than a couple of days to bring it back into full swing, you quickly realise the importance of a sound DR plan.
But implementing the right DR strategy for your business can prove tricky, unless you know what type of data you need to recover quickly to keep the business up and running efficiently.
The recent Symantec 2008 Disaster Recovery Research report, which surveyed 1000 respondents globally, found some alarming results. Of the one-third of organisations that had to execute a DR plan, just under half indicated they could be fully operational within one week. Another key finding showed 50 per cent of Australian organisations have had to put their DR plan into use either in full, or in part, because of a disaster. Fifty-two per cent were concerned about the cost of downtime and the reduction in revenue.
In the midst of a disaster
All types of organisations, whether big or small, face the same types of challenges when a disaster strikes. When data loss occurs, it has the potential to create a loss of revenue, impact productivity and customer relationships, and in some cases involve legal implications However, the difference between the challenges for small and large organisations is related to how much they intend to spend. For example, a big banking institution would most likely place a massive investment behind an efficient data recovery system, because it can’t afford a day’s worth of downtime.
A manufacturer, meanwhile, could possibly still keep functioning with only their most critical data assets while the rest of IT system plays a day or two of catch up.
But whatever the situation, one of the first steps organisations should take is to analyse what information is and isn’t critical for their business performance to determine the right type of data backup technology.
“Rather than looking at the primary technology, look at the information,” Gartner storage analyst, Phil Sargeant, said. “Organisations should understand their information a lot more intimately than they do. Then they can marry the appropriate technologies to the information.”
For instance, if some information needs to be recovered in minutes, rather than days, organisations could use disk-based systems for a quick recovery solution. But if the information is not that critical and it can be recovered over a couple of days, organisations won’t have to use such expensive mechanisms, Sargeant said.
Virtualising the environment could also cut 30-40 per cent off the cost of implementing a DR plan and therefore, should be considered for those needing a quick recovery.
“Once you virtualise, you’re not reliant on setting up exactly the same infrastructure on the other site as you did before,” EMC chief technology officer, Clive Gold said. “That’s a big change and a lot of customers usually start out by having a virtualised disaster recovery site, while they still have a completely physical main site.”
Virtualising the disaster recovery site is becoming one of the most popular ways for organisations to keep their data highly available. But in Australia, not everybody is convinced.
“Interestingly, 36 per cent of Australian companies are using tape, which is quite high when you think about the cost of disk, the technology advancements in disk-based backup has increased dramatically and the cost has decreased,” Symantec senior director of channel sales, David Dzienciol, said.
With or without data
Once you have analysed and categorised the information, there are other factors that can inhibit the level of preparedness, including limited resources and time.
“An SMB might not have access to appropriate skilled technical resources. Therefore, SMBs must leverage backup and recovery technologies to match their resources, yet meet their needs,” Acronis A/NZ general manger, Bill Taylor-Mountford, said. “If an SMB’s data protection solution does not address the complete lifecycle management of data, they risk unacceptable exposure of data that can easily result in loss and costly downtime.”
Additionally, every type of business environment contains different systems and each one provides a certain value. So figuring out the best type of technology to suit business requirements is not a task to be taken lightly. Identifying exactly what part of the company’s IT infrastructure is important enough to warrant a full-on recovery should also be part of any DR planning.
“Often IT departments are constrained by particular projects that may not have budgets to provide disaster recovery when they’ve been put in the first place,” Hitachi Data Systems chief technologist, Simon Elisha, said. “They’re often in the unenviable position of trying to provide a service with an appropriate budget or guidance.”
Some high-profile systems will be given all the protection necessary and will be set-up correctly, but supporting infrastructure that’s more techie in nature – yet fundamental to all the systems running – usually aren’t similarly protected.
“They can fail major systems across but they won’t operate without their ancillary systems as well,” Elisha said. “The first step of any disaster recovery process is the strategy around what you’re trying to achieve, the systems that are categorised as requiring disaster recovery and then giving the appropriate level of investment to enable them to be recovered. Most organisations in Australia don’t have all of their systems protected, only a percentage of them.”
Moreover, many organisations underestimate the amount of time it takes to fully recover from a disaster and don’t adequately identify what exactly needs to be recovered, Spectrum Data corporate data services manager, Dylan Gowrea, said.
“They might only backup certain systems or servers they feel are inadequate to be recovered. When it comes to the crunch they find themselves in a whole mess of problems.”
Road to recovery
And while disk backup has the greatest momentum as a technology when tackling these DR problems, tape hasn’t disappeared – it is being used more often as a final recovery point. But one of the biggest areas of growth has come from disk-to-disk or backup-to-disk-type solutions.
“Backing up to disk does help customers backup data a lot more quickly and, more importantly, get that data back quickly in the event of a failure. That’s not to say that tape is going to be the final tier, it just means that we’re running a two-stage backup,” HP StorageWorks South Pacific product marketing manager, Mark Nielsen, said.