Why OpenStack uses conductor

For reference.

Tan0 students to my explanation:

Two reasons one is to prepare for isolation because the upgrade is mainly the schema of the DB if the compute directly read and write to the DB, then each upgrade will have to ascend compute now isolated after the upgrade conductor, Then compute and conductor can communicate through an object.

The second reason is that SQLAlchemy read the DB will block live if let compute to operate slqalchemy the words will appear problems, now compute can throw the task to conductor, you can go to perform other tasks waiting for the result is better
Cloud-compliant parallel architecture

Why OpenStack uses conductor