Implement `zero downtime` by message queuing.

[kaxada]

Is your feature request related to a problem? Please describe.
Currently when there is a new deployment/build, the old docker container is deleted and the process creates a downtime which fails some requests from processing.

Describe the solution you'd like
The solution is to implement a message queue that handles and monitors the success of requests separately. this will be done with RabbitMQ

Describe alternatives you've considered
I had also considered load balancing which is a faster approach but would require some extra resources but didn't look into into extensively.