In the previous section we discussed the first part of Cinder creating Volume, Cinder-api operation, this section continues the second part, Cinder-scheduler scheduling work.
cinder-scheduler Execution Scheduling
Cinder-scheduler executes the scheduling algorithm and selects the optimal storage node log for/opt/stack/logs/c-sch.log through Filter and Weigher.
Cinder-scheduler performs the dispatch work through Flow Volume_create_scheduler.
The Flow executes Extractschedulerspectask and schedulecreatevolumetask in turn.
The main filter and weighting work is done by Schedulecreatevolumetask.
After Availabilityzonefilter, Capacityfilter, Capabilitiesfilter and capacityweigher layers of filtering, the storage node was finally selected [email protected]# Lvmdriver-1.
Flow Volume_create_scheduler completes the dispatch and the status changes to SUCCESS.
Cinder-scheduler sending Messages
Cinder-scheduler send a message to Cinder-volume, let it create volume source/opt/stack/cinder/cinder/scheduler/filter_scheduler.py, method for Schedule_create_volume.
In the next section we discuss the last part of Create Volume: the cinder-volume process.
Create Volume operation (part II)-Play with OpenStack (51) 5 minutes a day