Multi-teenage architecture: definition
If there are different types of clouds (public, private, hybrid), there are also several categories of architecture, the multi-tenant of which is part.
To understand this, we must first define the term “holder”, which can be translated as “tenant” in French. In most cases, this tenant is a client (a company) made up of several users (employees).
Thus, in a Multi-benefit infrastructure (or multi-thirds)the same instance of a software application is used by several customers isolated from each other. Software with such an architecture has a unique technical installation, shared by all users.
The multi-tenant mode is characteristic of SaaS applications, as opposed to the on-premise model, in which each company installs its version of the software on its server. It is also distinguished from single-teenage architecturewhere each customer is the only user of its software and has its own database and its own application server.
The characteristics of multi-teenage architecture
Insulation of data and execution environments plays a key role. As such, several approaches are possible to separate data from the various tenants.
Distinct databases
This approach is to Store customer data in different databases. It simplifies the extension of the application data model to meet specific needs, offers a high level of safety and allows fast food of the data of a “holder” in the event of a breakdown.
Shared database and separate patterns
Here, several tenants are hosted in the same database. Each of them has a set of tables, grouped in a diagram created specifically for the customer.
Flexible and easy to implement, this approach, however, offers lower security guarantees, as the system is not completely isolated. In addition, the restoration of a tenant’s data can be more complex than with separate databases.
Shared database and diagram
In this approach, All customers use the same database and the same diagram. It generates several additional constraints: risk of bugs linked to overlapping recordings, higher test costs, drop in performance due to the accommodation of all recordings in the same table …
The advantages of multi-teenage architecture
In general, the multi-tenant model offers several advantages vis-à-vis other types of architectures.
Rapid deployment
This model simplifies the addition of new customers and reduces the manual work necessary for the implementation of the solution. This results in a significant increase in Time-to-Valuethat is to say the speed with which customers can benefit from the value brought by the application.
Reduction of costs
The use of a common infrastructure results in a lower costs, because the resources are shared between the different customers. That’s why Prices for multi-teenage SaaS solutions are generally more attractive.
In addition, multi-third architecture reduces the total cost of possession: An additional user costs almost nothing. Conversely, in a single-tenant system, the addition of a new customer often represents a very high cost, which includes the acquisition of the new equipment, but also the costs related to installation, configuration, maintenance or upgrades.
Practical maintenance
A multi-teenage application is much more practical and rapid to maintainsince it is based on a single infrastructure. When a modification is made to a file, it immediately benefits all customers. In addition, the publisher can effectively manage many customers from a single interface.
Flexibility in the event of a rise in load
In case of a rise in charge, Multi-teenage architecture makes it possible to easily allocate additional resources to an application in progress.
In other words, if software requires a number of servers to manage an additional load and the user faces a traffic peak, it is possible to add more capacities in just a few seconds.
Simplified updates
Multi-teenage architecture considerably simplifies the update process.
Indeed, since it is based on a single infrastructure, updates are less restrictive and can be performed more regularly. Just apply them only once so that all users take advantage of them.
With each version change, the software is automatically modified. Customers therefore do not need to make a long and tedious update on their own server.
Shared data analysis
Finally, software with multi-teenage architecture facilitate the use of analysis tools, such as dashboards and reports, for all customers.
Multi-teenage and multi-instance: what differences ?
The multi-teenant should not be confused with Multi-instance architecturein which each customer performs their own application of the application, with its own database.
It guarantees a Total customer data insulationbecause each tenant can access his data separately from the others. This results in a high level of security and confidentiality for user companies. The overall availability of the infrastructure is also very high: in the event of a breakdown of an instance, the problem will only affect one customer.
In addition, this model quickly increases a customer’s resources, because only its infrastructure must be modified. It is therefore possible to allocate additional capacities according to his needs, with a simplified scaling. The possibilities of personalization, with, for example, the addition of dedicated features, are also very numerous.
Nevertheless, multi-instance architecture is more complex to deploy and maintainsince each customer has their own dedicated infrastructure. Each of them must be deployed, maintained and updated separately. In comparison, the implementation of a multi-tenant architecture is much easier.
Finally, Multi-outs generates higher costsbecause they are not shared. Thus, the creation and configuration of an application or database is less profitable.
