Storage and the associated planning process are critical for an organization to consider any cloud technology. In this TechRepublic blog post, IT Jedi Rick Vanover outlines one solution that may be attractive for availability.
I will admit that cloud technologies are a buzzword. Primarily because there are many different types of clouds and how one organization perceives cloud technology can be fundamentally different than another. In last week's post on cloud storage, I mentioned how the Amazon Web Services (AWS) storage offerings are provided. For the Simple Storage Service (S3) cloud, availability is described on their Web site as being 99.99% available. Comparatively the Elastic Block Store (EBS) cloud is advertised as highly available -– yet without a percentage like the S3 cloud. The EBS service offers an automatic replication feature, which is different than the S3 cloud. This automatic replication replicates allocated EBS volumes to the same zone.
That is all fair and well until either you fall into the .01% of the time that it is unavailable or a zone of the EBS cloud is not accessible. That is where an interesting technology was made available to me on highly available cloud storage. Recently, I attended a CloudCamp event that was a great resource to see what people are doing and to understand the technologies in play. Sure, there are plenty of sales opportunities at these events. What I found more interesting was access to how these technologies work as well as what organizations are already there.
One offering that caught my eye was the Nirvanix Storage Delivery Network (SDN). The SDN can provide a 100% service license agreement (SLA) for certain configurations and data sizes. This would be a scenario where you have 5 TB of data or more in the SDN cloud and are selecting a data replication policy that covers 3 nodes (different datacenters in the Nirvanix namespace).
The underlying prerequisite, of course, is that the customer can access the data. So, strategies of WAN failover or forcing all workers to telecommute can come into play.
Availability in the cloud is a contentious topic, as the TechRepublic community demonstrates. I am providing cloud resources so we can understand the technologies -– I'm not necessarily saying we should go there. I am a bit of a "server-hugger," I will admit that. But I see cloud as the single biggest threat to internal IT infrastructure on the radar right now. Share your comments below.
Rick Vanover is a software strategy specialist for Veeam Software, based in Columbus, Ohio. Rick has years of IT experience and focuses on virtualization, Windows-based server administration, and system hardware.