Multiple Workspace vs Separate Guzzle instances - ja-guzzle/guzzle

Multiple Workspace

Leverage same node and API instance reducing overheads ,but on flip side there is scalibility issue with this option
Ability to allow aggregated view of jobs runs
Introduces two level of configs - the instance level and workspace level
May reach scalability issue as Guzzle offers more services like Scheduling and Rest API specially when more workspaces running on same instances

Clear isolation of each workspaces which run in separate instances - keep security deign simple, keeps the admin and code path simple
We can scale each instance independently
Simplifies manageability including provision of new instance
Metering is very clean for the Guzzle VM
In future the instances can run on docker- reducing overheads of OS as customer try to host more instances
Simplifies some of the global guzzle concepts like: guzzle.yml where we store context columns
Does not introduce two levels of config - its just one now
Aligns to how Guzzle was design from day1.

To go with separate instances for incases where customer want to use Guzzle for different application
Also separate env will be separate instances
Most of the cloud services are designed to run independent instances this applies to: Data Factories, DB Workspace etc.