In the micro service architecture
Parsing, large files is now four analytical services, a distribution service:

1, through the distribution of evenly distributed file service number to each analytical services;
2, the four analytical services at the same time;
3, because of the need to parse the file size to differ, to more small file distribution services (assuming that two) is done soon, is idle; Assigned to the more large file distribution services (assuming that two) is still in the working state;
4, due to the total time consumed by a need to get the whole system, the two have complete free services are idle, but on the other two working status of the service in working state;
So when parsing the file is too large, two work service has been time consuming, and two free service does not work, it is a waste;
5, how to fulfill the task ahead of time of free services, to help also worked in the service to share the pressure of work???????

CodePudding user response:

Dear, have some idea

CodePudding user response:

I think you have to change the way of thinking,
Optimization cannot be after the distribution, and should be on assignment before!

CodePudding user response:

refer to the second floor paullbm response:

I think you have to change train of thought,
Optimization cannot be after the distribution, and should be on assignment before!

Uh-huh, I also have been thinking about it this way