Multitenancy dispatcher best practice

[MatthewDa19]

Hello all,
I know in a multi-tenant AEMaaCS setup, each tenant’s code can be included as a Git submodule within a main repository, and dispatcher configurations can be customized per tenant. However, I’m looking for best practices: How can I manage tenant-specific dispatcher configurations (like cache rules, filters, and vhosts) to ensure proper isolation between tenants while maintaining a shared deployment pipeline?

[SantoshSai · Accepted solution]

Hi @MatthewDa19,

You may need custom cache rules, filters, or vhosts for each tenant Since dispatcher configs are shared, the best practice is:

1. Centralized dispatcher configuration repository:

   • Maintain a dedicated Git repository for all dispatcher configs.

   • Include tenant-specific configs as separate directories or submodules.

2. Modular dispatcher configs:

   • Store each tenant’s config in separate files and include them via $include.

     /farms
     {
         $include "farm-tenantA.any"
         $include "farm-tenantB.any"
         $include "farm-common.any"
     }

3. Automation & environment management:

   • Use CI/CD scripts to deploy tenant-specific configs automatically.

   • Maintain separate configs for dev, stage, and prod; use environment variables to adjust paths dynamically.

Reference for more details:

• Basic Multi-Tenant Model for AEMaaCS – Perficient