Combined cluster & tenancy architecure for workflow system

[pahab]

We are currently trying to figure out how the architecure of our workflow system should be. We are considering something like „cluster process engine installation with multi tenancy“:

• One database for all tenants
• Clustered Process Engines with multiple tenants

Basically, a combination of these two architecture images:

and

Clustered Process Engine should get tasks from the shared database via tenant ids. However, in the documentation and the examples found, the multi tenancy installation had a process engine per own tenant setup.

Is it possible to have a cluster setup with multiple process engines combined with the multi tenancy in one database?
In this setup can each process engine work for each tenant?
And can a process/taks be bound to a specific process engine?

I appreciate your help

[Philipp_Ossler]

Hi @pahab,

it’s possible to use multi-tenancy also in a clustered setup. If you use one shared database with only one schema for all tenants then every process engine in the cluster can access all data.

That means, a process / a task isn’t bounded to a specific process engine because you use one schema for all tenants - in contrast to ‘one-schema-per-tenant’ concept.

If you want to limit the access to specific tenants then you have to set the tenant identifiers when you query the data. See the user guide for details.

Does this answer your questions?

Best regards,
Philipp