From Eager Futures/Promises to Lazy Continuations Evolving an Actor Library Based on Lessons Learned from Large-Scale Deployments Benjamin Hindman @benh CppCon 2021prologue ● past life at UC motivating futures/promises + actors (2) libprocess (3) revisiting the problem (4) evolution of libprocess (5) eventuals (6) scheduling (7) streams (8) type erasurechapters (1) motivating futures/promises different language (e.g., Erlang) … or bring Erlang to C++! use callbacks … more on this later use futures/promises …possible solutions just wait … use threads … too expensive, not conducive to correctness

automatically by pip or setup.py install, the following optional packages may be useful: • concurrent.futures is the recommended thread pool for use with Tornado and enables the use of ThreadedResolver. It my_future.set_result(f.result())) return my_future The raw Future version is more complex, but Futures are nonetheless recommended practice in Tornado because they have two major advantages. Error handling exception (as opposed to the ad-hoc error handling common in callback-oriented interfaces), and Futures lend themselves well to use with coroutines. Coroutines will be discussed in depth in the next section

in Tornado generally return placeholder objects (Futures), with the exception of some low-level components like the IOLoop that use callbacks. Futures are usually transformed into their result with the exceptions have somewhere to go. • Decorated coroutines have additional integration with the concurrent.futures package, allowing the re- sult of executor.submit to be yielded directly. For native coroutines, way to call a blocking function from a coroutine is to use IOLoop.run_in_executor, which returns Futures that are compatible with coroutines: async def call_blocking(): await IOLoop.current().run_in_executor(None

in Tornado generally return placeholder objects (Futures), with the exception of some low-level components like the IOLoop that use callbacks. Futures are usually transformed into their result with the have somewhere to go. • Decorated coroutines: – have additional integration with the concurrent.futures package, allowing the result of executor.submit to be yielded directly. For native coroutines, use way to call a blocking function from a coroutine is to use IOLoop.run_in_executor, which returns Futures that are compatible with coroutines: async def call_blocking(): await IOLoop.current().run_in_executor(None

in Tornado generally return placeholder objects (Futures), with the exception of some low-level components like the IOLoop that use callbacks. Futures are usually transformed into their result with the have somewhere to go. • Decorated coroutines: – have additional integration with the concurrent.futures package, allowing the result of executor.submit to be yielded directly. For native coroutines, use way to call a blocking function from a coroutine is to use IOLoop.run_in_executor, which returns Futures that are compatible with coroutines: async def call_blocking(): await IOLoop.current().run_in_executor(None

in Tornado generally return placeholder objects (Futures), with the exception of some low- level components like the IOLoop that use callbacks. Futures are usually transformed into their result with the have somewhere to go. • Decorated coroutines: – have additional integration with the concurrent.futures package, allowing the result of executor. submit to be yielded directly. For native coroutines, use way to call a blocking function from a coroutine is to use IOLoop.run_in_executor, which returns Futures that are compatible with coroutines: async def call_blocking(): await IOLoop.current().run_in_executor(None

in Tornado generally return placeholder objects (Futures), with the exception of some low- level components like the IOLoop that use callbacks. Futures are usually transformed into their result with the have somewhere to go. • Decorated coroutines: – have additional integration with the concurrent.futures package, allowing the result of executor. submit to be yielded directly. For native coroutines, use way to call a blocking function from a coroutine is to use IOLoop.run_in_executor, which returns Futures that are compatible with coroutines: async def call_blocking(): await IOLoop.current().run_in_executor(None

in Tornado generally return placeholder objects (Futures), with the exception of some low- level components like the IOLoop that use callbacks. Futures are usually transformed into their result with the have somewhere to go. • Decorated coroutines: – have additional integration with the concurrent.futures package, allowing the result of executor. submit to be yielded directly. For native coroutines, use way to call a blocking function from a coroutine is to use IOLoop.run_in_executor, which returns Futures that are compatible with coroutines: async def call_blocking(): await IOLoop.current().run_in_executor(None

in Tornado generally return placeholder objects (Futures), with the exception of some low- level components like the IOLoop that use callbacks. Futures are usually transformed into their result with the have somewhere to go. • Decorated coroutines: – have additional integration with the concurrent.futures package, allowing the result of executor. submit to be yielded directly. For native coroutines, use way to call a blocking function from a coroutine is to use IOLoop.run_in_executor, which returns Futures that are compatible with coroutines: async def call_blocking(): await IOLoop.current().run_in_executor(None

in Tornado generally return placeholder objects (Futures), with the exception of some low- level components like the IOLoop that use callbacks. Futures are usually transformed into their result with the have somewhere to go. • Decorated coroutines: – have additional integration with the concurrent.futures package, allowing the result of executor. submit to be yielded directly. For native coroutines, use way to call a blocking function from a coroutine is to use IOLoop.run_in_executor, which returns Futures that are compatible with coroutines: async def call_blocking(): await IOLoop.current().run_in_executor(None