Just October last year, the United States was hit by super storm Sandy – it brought widespread flooding and power outages on the north-eastern US seaboard. With hyperscale cloud providers having majority of their data centers in the US, most businesses would want to ask: “Can natural disasters kill the cloud”?
The hurricane which hit US this October brought cloud providers into an immediate disaster recovery mode but still caused downtimes for its B2B markets. For instance, web publishers such as Huffington Post, Gawker and BuzzFeed suffered outages as Datagram (its cloud provider experienced flooding and fibre outages in its New York and Connecticut datacentres). Verizon‘s data centres were also downed, and its generators failed to pump fuel due to flooded basements. As an attempt to mitigate the risks brought by Sandy, Nirvanix offered its New Jersey customers to have their data moved to Frankfurt or Tokyo either temporarily or permanently with no charge to the clients.
The Cloud Risks Brought by Natural Disasters
This hard reality states that although cloud providers implement the highest security measures and put multiple data centers in diverse locations on-shore and off-shore to perform data redundancy and disaster recovery; downtimes, the potential for data loss and even security breach is still probable. With these risks on customers’ plate, what is the most logical step to do? Should businesses just opt to have their datacentres on-premises; back-up data using tapes, external hard disks, like they did before cloud?
Natural disasters can bring down a datacentre, this is a fact. However, the rate in which cloud providers act in mitigating such risks makes the difference. In effect, although it makes businesses feel at ease bn the notion of having a physical backup of their data in their respective on-premises datacentres, this is not only an expensive approach but also makes them more prone to data loss and outages.
If a natural disaster hits an on-premises datacentre and all of the backups start and end with the same servers and office, chances are there will be little left to save after the turmoil. The cloud however performs data redundancy, fail-over mechanisms and provides SLAs that safeguard data against natural disaster. The difference now is actually not just the technology, rather on the type of cloud service provider.
The Microsoft Advantage
For example, Windows Azure SQL Database – one of Microsoft’s cloud offerings – maintains three replicas of each database: one primary and two secondary replicas. All pushed data on Azure’s SQL Database are replicated to its mirrored instances. In the event that the primary replica fails, Azure will automatically detect the issue and fail over the secondary replica, thus preventing not only loss of data but also potential downtime. Other than the backup database, Azure’s database fabric can maintain a minimum of 14 days worth of backups taken in five minute increments. These backups are archived in different local datacenters to guard against hardware or system failures.
Other than the built-in measures that Azure has, it also empowers its users by allowing them to create database copies. The backup process is asynchronous; in effect this can be executed through a simple copy command and have its progress monitored. Other than maintaining database copies, customers can also use the SQL Database Import/Export Service – this enables to pull and push data and schema in a package that can contain tables, views, indexes, constraints, triggers, stored procedures, logins, users etc.
No technology is perfect, including the cloud. Although providers are doing their own share by implementing various safeguarding measures, they are not equal. Truth is some providers handle risks like Sandy better than their counterparts.
When choosing the right cloud vendor, make sure to look at these aspects.
Discover more from Loryan Strant, Microsoft 365 MVP
Subscribe to get the latest posts sent to your email.