all of the activity in C5 happens on a combination of fibers (jetlang construct) and raw threads (various threadpools, etc: in netty pools). Handoff has to be managed carefully, and we want to manage and be aware of how much blocking IO is being done on fibers.
One problem facing this is that the information to understand this is spread out in the code base. Centralizing it and curating it and documenting it would be nice.
all of the activity in C5 happens on a combination of fibers (jetlang construct) and raw threads (various threadpools, etc: in netty pools). Handoff has to be managed carefully, and we want to manage and be aware of how much blocking IO is being done on fibers.
One problem facing this is that the information to understand this is spread out in the code base. Centralizing it and curating it and documenting it would be nice.