# different job_serializer/job_parser function by tube

I using Beanstalk as primary message queue with Beaneater. In my scenario, every service got its own Beanstalkd Tube.

As system growing, There is an idea about using optimized job serializer/parser for different service to get better performance. 

Any suggestion  :D?