Making Databases Elastic in the Cloud

Michael Vizard

As cloud computing evolves, IT organizations are finding that the biggest challenges they face are finding ways to scale databases across the cloud.

The vast majority of the databases installed today were never intended to dynamically scale. To make these systems work in the context, database administrators have had to 'shard' the database to make it scale. The problem with database shards is that not only are they difficult to get right, but they are hard to manage on an ongoing basis.

The folks at Clustrix, a provider of clustered databases that are deployed on an appliance, think that a better approach would to create a MySQL-compatible database platform that can scale by dynamically adding new nodes. The database automatically discovers each node and then redistributes data to optimize performance based on the policies defined by the IT organization. Dan Liddle, head of marketing and business development for Clustrix, says this approach eliminates the need for sharding the database.

One of the tenets of cloud computing is that data management needs to be elastic. But if the underlying data management system isn't capable of truly supporting that, what you wind up with is a lot of cloud-ready IT infrastructure that doesn't have all that much to do, especially when it comes to running mission-critical applications.

It's still the early days of cloud computing, so most IT organizations have not begun to think through their overall data management strategies in the cloud. And yet, when it comes to cloud computing, data management is the single biggest impediment to adoption.

Given all the financial pressures to adopt cloud computing, IT organizations are going to be under a lot of pressure in 2011 to start solving these issues. And more than likely, that may require a new approach to building and maintaining the underlying data management system.

Add Comment      Leave a comment on this blog post
Feb 9, 2011 4:02 PM Avi Kapuya Avi Kapuya  says:
License costs for this solution is very high comparing to other, pure software based solutions. Xeround Cloud Database offers high level of scalability and high availability, and a pay per use pricing model. Generally, Appliances are not really native cloud components, and in fact are contradicting with current cloud trend, using commodity hardware for all needs. on the blog post on native cloud services this issue is discussed more in depth: Reply
Feb 10, 2011 6:02 AM inventory pos sytsem inventory pos sytsem  says:
I appreciate your post, thanks for sharing the post, i would like to hear more about this in future. Reply
Feb 10, 2011 9:02 AM Razi Sharir Razi Sharir  says:
DB elasticity is a key requirement on a Cloud DB solution. This will often get translated to a barrier when looking at scaling out (adding nodes) vs. scaling up, and not limited to the management only but rather to all the layers in the architecture. We see many attempts to address this issue that claim elasticity but in fact can only show linearity within a single node which affects availability as well when one needs to add a node or repartition data� At Xeround, we have solved this featuring a s/w only solution as already demonstrated in our public MySQL DBaaS beta ( Reply
Jun 7, 2011 7:06 AM Office 2007 Office 2007  says:
And more than likely, that may require a new approach to building and maintaining the underlying data management system. Reply

Post a comment





(Maximum characters: 1200). You have 1200 characters left.




Subscribe to our Newsletters

Sign up now and get the best business technology insights direct to your inbox.