Data Backup Digest

The term “software-as-a-service” has been thrown around a lot as of late, but it’s not just some industry buzzword. Software-as-a-service, which saves enterprises the cost of having to establish and maintain their own IT infrastructure, has a number of real implications and benefits in the 21st century. One such benefit gives service providers the ability to host multiple customers within a single physical or virtual environment.

Known as multi-tenant database architecture, it effectively lets customers, or tenants, share the same database application. Although it is shared with another customer, tenants who are placed on a multi-tenant database will still retain the ability to customize and configure their own infrastructure to meet their needs and specifications.

Types of Multi-Tenant Database Architectures

There are currently three generic types of multi-tenant database architectures that are commonly seen today. Isolated tenancy sees the least amount of resource sharing as customers are given access to separate applications and separate databases, though some resources are still shared between tenants.

Infrastructure tenancy affords separate applications and separate databases, but multiple tenants are still sharing the general infrastructure. This multi-tenant option establishes a set of tables and schema that is specific to each tenant within the database.

The third and final type of multi-tenant database, known as application tenancy, sees shared applications and infrastructure in tandem with a separate database. All three options feature integrated database security ensures that tenants cannot access or modify the data of others.

Pros & Cons of a Multi-Tenant Database Architecture

As mentioned earlier, there are a number of ways a multi-tenant database architecture can benefit service providers of today. For starters, the introduction and installation of software updates is much easier, considering that less software instances need to be updated. Not only does this save on costs, but it also cuts down on the amount of time needed to perform such tasks on a regular or repeated basis.

Data aggregation and data mining efforts are also enhanced. Instead of gathering data from various sources, service providers are able to perform queries involving numerous customers and different datasets. As such, the processes of data mining and trend analysis are greatly simplified.

However, there is a downside to the multi-tenant database architecture. Customizations and modifications must be made in order to uphold individual tenant brands, to accommodate varying workflows and to provide each client the ability to manage access privileges for each of their own users. Moreover, the ongoing maintenance of per-tenant metadata calls for increased network complexity as well as the allocation of more resources for development purposes.

IBM cited some of the pros and cons of multi-tenant database architectures in a recent whitepaper, which read: “In a multi-tenant environment, a SaaS company can reduce costs if it shares or reuses more of its resources. However, the more the company shares resources, the more risks it faces because an outage of a shared resource can potentially affect many customers. The more resources shared also adds to the complexity of the solution.”