Multi-Tenancy: An Old Idea Made New Again in the Cloud
When big, room-sized mainframe computers emerged in the 1950s, they cost so much that making efficient use of them was at a premium. Machines with no computing power of their own (called “thin clients”) were used to tap into the mainframe’s computing power simultaneously.
The more things change, the more they stay the same.
The same concept underlying the old idea of “time sharing” is at work today in what’s called “multi-tenancy.” In the context of cloud computing, this refers to sharing one instance of a software application.
In other words, several customers (called “tenants”) use the same software application on the same hardware and operating system, but don’t see each other’s data or what other customers are doing. So rather than having, say, individual desktop computers loaded with separate email software, the cloud provider can make use of one instance of that email software, allowing many people to use it at the same time.
Those users will never see each other’s email, but the cloud provider will be able to upgrade all of their email software in one fell swoop, rather than having to go each person’s desktop to do the job.
This approach also makes more efficient use of servers and other technology and cuts costs. To wit:
- Using one instance of a software application reduces the use of memory and processing power. Also, one seat license for the software is all that’s necessary, rather than purchasing multiple licenses for a group of users.
- All the data for all users is stored in one spot, making it easy to find trends in use and similar tidbits.
- One installation of code for alterations to the software application is all that’s necessary.
Forrester’s Findings on Multi-Tenancy
Forrester Research, a market research and consulting shop of which Connectloud is a client, has studied multi-tenancy and made three key findings:
- Cloud operators must strike a balance between the sharing of an application and the customers’ security. In other words, a multi-tenant operation must balance the differing needs for computing resources, on the one hand, with each customer’s own need to have their data kept private and secure. “These two goals ultimately conflict with one another, since shared resources and individual security rarely go hand in hand,” Forrester says.
- Forrester outlines two separate multi-tenant models. One, called “Dedicated Resource” models, essentially draws a virtual line in the sand about what technology resources a tenant can access and what it can’t. That feels more secure, but has less flexibility for both tenant and operator. The other model, “Metadata map,” offers more flexibility but may feel less secure, Forrester says.
- A multi-tenant architecture will generally provide more security even though it involves sharing computing resources. “Multi-tenant services secure all assets at all times, since those within the main perimeter are all different clients,” Forrester says.
uCloud is About Security, Agility, Scalability, Cost-Effectiveness
When you break it down, Connectloud’s uCloudTM platform was built around the balance of agility, cost-effectiveness and security to which Forrester refers. Our mission is to deliver a unified cloud solution that is simple and quickly scalable multi-tenant automation with reduced operating expenditures.
We understand that the needs of enterprise today extend beyond singular data clouds. The need now is to manage and automate across clouds. Thus, uCloud – a next-generation cloud on top of multiple clouds, whether public, hybrid or private.
Or, you can think of it as we do: True business value.