Software As A Service

Multi tenancy and SaaS terms are using interchangeably, to represent different point of views for same architecture approach that target providing services for multiple customers on the same architecture and database. 

SaaS: software as a service, a new approach that impacts the solution architecture model that enables software organizations to build solutions that serve more customers instead of one customer, which improves their pricing model, and moves from license to rent the solution as a service, using architecture techniques for both SW and HW.

 The opposite graph shows the difference between traditional computing and cloud computing, the base for SaaS.

So, Multi tenancy means: 

• shared services and resources: which include Compute, network, and storage 

• For multiple customers

Software Companies:

• Cost

  • • Higher Cost

• Revenue

  • • Competitive price

    • Higher revenue

Customer Perspectives: 

• Zero fixed cost

• Lower operation cost

• Fast kick-off

• Zero software license

• Business Continuity

We have three architecture models for the database:

1. Database per tenant

2. Schema per tenant

3. Shared database

Your application will have a separate database for each tenant, which we can mention the pros and cons as following:

• Pros:

  • • Best isolation 

    • High performance 

    • High security

    • Ease backup

• Cons

  • • Expensive

    • Deployment cost

    • Low maintainability

• Pros:

  • • Better Isolation

    • Easier maintainability per schema

• Cons

  • • Complicated when scaling up to a microservice architecture

    • Hard deployment process

• Pros:

  • • Low cost

    • Ease of scalability

    • Added feature for shared tenancies and private tenancies can take place to meet governmental requirements and security policies, for example:

      • • Small entities can live with shared tenants as they focus on cost

        • The Ministry of Defense needs a dedicated tenant 

• Cons

  • • Complex Query filtering in shared tenants

    • Potential data leakage, which impacts security and privacy

• Application of a VPD: virtual private database to take place

• Use row-level security (RLS) for the shared schema.

• Enforce strict tenancy boundaries at the service/API layer.

• Design with database provisioning automation for onboarding/offboarding.

• Encrypt data at rest and in transit.

• Use tenant-aware indexing and sharding if needed.

• Apply DevSecOps principles and audit logging.