Downtime can be fatal in the digital age, warns Veeam

'Nine out of ten companies that have suffered a serious data centre outage have been forced to close operations,' says Russell Nolan

Enterprise architecture is becoming more complicated, spread across data centres and various cloud providers. At the same time, the requirement for 365x24x7 uptime and robust data security is growing more pressing by the day.

The capabilities offered by cloud have ushered in the digital era, raising expectations accordingly. So now when Office 365 goes down for an hour, or more seriously a VM goes offline in an airline's data centre, the consequences can be severe, as BA found out recently.

Entire business processes are now predicated on systems being always-on. The advent of multi-cloud or best-in-breed cloud services brings new possibilities to large organisations, but at the price of yet more complexity. There's an old adage that complexity is the enemy of security, so how can the imperatives of increased functionality and availability be matched with those of data security and compliance?

The answer, in part at least, is a focus on the basics, stored on two different media with at least one being maintained offsite.

"Companies should adhere to best practice like the 3-2-1 rule," Russell Nolan, Senior System Engineer at data management firm Veaam told the audience at Computing's Cloud & Infrastructure Summit last week. The numbers, Nolan explained, stand for three copies of all important files, saved on two different media with one being offsite.

Among most popular uses of public cloud is BaaS or backup as a service, so one might have assumed that the ‘one' aspect of the 3-2-1 rule - offsite storage - would be pretty much taken care of by now.

Not so, Nolan said.

"Almost 40 per cent of businesses don't send their data offsite," he said. "That means that they are not following the 3-2-1 rule".

The reasons for this are around costs and complexity, he said. It is not just about securing data, it's being able to recover it at a moment's notice.

"Nine out of ten companies that have suffered a serious data centre outage have been forced to close operations," Nolan pointed out.

While many individual cloud providers may detail recompense for any outages in their SLAs, this will often arrive after the damage has been done. Nolan gave the example of Office 365.

"Microsoft's SLA is only 90 days," said Nolan "It's enforced financially, so they will pay you compensation, but if your email goes down with 89 days to go that's not particularly helpful."

Applications hosted in cloud IaaS like AWS or Azure are also vulnerable to service outages. Again, compensation may be available, but it could be too little and too late to make up for damaged reputation and financial loss.

With data and applications spread over so many different platforms, managing backup, governance and recovery requires an overarching platform that can centralise the management of backup and recovery, Nolan said.

As a company Veeam, has grown from its origins in data centre backups, to providing backups, recovery and availability in the hybrid cloud - now the de facto operating model in many organisations. However, there has historically been an imbalance between the controls available across the various platforms.

"Providing granular access in the public cloud space has always been difficult," Nolan explained.

However, this is now changing and Veeam's agentless solutions are able to provide backup and availability across AWS with agent-based software doing the same for data and applications hosted on Windows and Azure.

The company's soon-to-be-released Availability Orchestrator aim's to tie all this functionality together by providing enterprises with a DR orchestration engine for the Veeam's backup and recovery and availability suites.

"It does more than just back things up, it provides more control over hardware changes during failover and failback and it has built-in documenting, updating, and reporting of DR plans for compliance," Nolan said.