![Receiving messages
Rubinius
def actor_process()
looping = true
while looping
Actor.receive do |鍖lter|
鍖lter.when Hello do |message|
greet(message.from)
end
鍖lter.when Goodbye do |message|
message.from << KThxBye[self]
looping = false
end
end
end
45](https://image.slidesharecdn.com/comparingimplementationsoftheactormodel-130216225753-phpapp01/85/Comparing-implementations-of-the-actor-model-45-320.jpg)
Comparing implementations of the actor model
- 1. Concurrency Summit 2008 Comparing implementations of the Actor Model Jim Roepcke <jr@uvic.ca> 1
- 2. Actor models? 2
- 3. This kind? 3
- 4. No, not that kind... 4
- 5. The kind in Joe Armstrongs Erlang, 5
- 6. The kind in Joe Armstrongs Erlang, 6
- 7. Evan Phoenixs Rubinius, 7
- 8. Martin Oderskys Scala, 8
- 9. and Steve Dekortes Io. 9
- 10. What is the Actor Model? 10
- 11. Its a trendy model for concurrent programming 11
- 12. Massively parallel future The future has begun! We had lots of time to 鍖gure this out Programmers are still using threads and mutexes in high level code Bad programmer! ABSTRACTION FTW! 12
- 13. Whats old is new again Actors de鍖ned by Hewitt in 1973 More work by his PhD students in the 80s, Clinger, Agha Adopted by Erlang in the late 80s for the telecom industry Picking up steam as concurrency becomes a hot issue again 13
- 14. Actors win because they are simple 14
- 15. Actors communicate 15
- 16. Actors spawn 16
- 17. Actors run in parallel 17
- 18. Actors listen 18
- 19. Actors avoid deadlock 19
- 20. Actors perform great! 20
- 21. Now: up close 21
- 22. Comparing Actors to Actors Implemented a program in Erlang, Rubinius, Scala and Io My experiences with each 22
- 23. OM NOM NOM! Implemented the burger problem from CSC 464 Assignment 1 in all four languages 23
- 24. spawn! supervisor 24
- 25. mmm hungry Student Student Student Burger burger! Student Student enjoy! Order Line Burger Line ok k Cook Cook Burger Burger supervisor 25
- 26. Actors communicate 26
- 27. Communication Asynchronous Message Passing Encourages no memory sharing Each actor has a mailbox which queues messages sent send mechanism to deliver a message to an actor 27
- 28. Sending messages Erlang Rubinius pid ! message actor << message Scala Io actor ! message obj @@method(args) 28
- 29. Actors and burgers 29
- 30. Actors spawn 30
- 31. Creating actors To create concurrency, actors create more actors and send them messages Actors are independently executing entities 31
- 32. Creating actors Erlang Rubinius pid = spawn(fun) actor = Actor.spawn ... Scala Io sa = new SomeActor() sa := SomeActor clone sa.start() 32
- 33. Erlang Actors Actors are called processes Process creation and context switching is very fast and lightweight Processes cannot share memory Typically run a tail-recursive function 33
- 34. Rubinius Actors Actor class comes with Rubinius VM Green Ruby thread per actor VM assigns a Mailbox per thread Remote Actors new, not perfect yet Orders of magnitude worse than Erlang: Rubinius developer on IRC 34
- 35. Scala Actors scala.actors package comes with Scala Hybrid execution model, actors run in their own thread or via events Claim massive scalability using events and thread pools Remote Actors supported act() loops to stay alive 35
- 36. Io Actors Io uses coroutines for concurrency Scheduling is FIFO, no prioritization One kernel thread per VM, async I/O @@ runs the method on a dedicated per-object coroutine for messages Remote actors not supported 36
- 37. Actors run in parallel 37
- 38. Parallelism: Erlang SMP Erlang automatically puts processes on different cores/CPUs Messaging actors in remote VMs is seamless spawn(Node, fun) to spawn an actor on a different VM which may be remote 38
- 39. Parallelism: Rubinius One kernel thread per Rubinius VM Processor af鍖nity for multi-core? Use VMActor to create an actor on a remote VM Messaging actors in remote VMs is seamless thanks to duck typing 39
- 40. Parallelism: Scala When using event-based actors, Scala creates multple threads to balance load across cores Use RemoteActor class to 鍖nd an actor on a different VM Messaging actors in remote VMs is seamless 40
- 41. Parallelism: Io Nope :-( One kernel thread per VM No support for remote actors built-in 41
- 42. Actors listen 42
- 43. Receiving messages Take a message off the mailbox queue Decide if you want to deal with it Messages are data Timeouts 43
- 44. Receiving messages Erlang loop(State) -> % a tail-recursive function receive {hello, From} -> greet(From), loop(State) ping -> doPingThing(), loop(State) {goodbye, From} -> From ! {kthxbye, self(), State} % no loop after 100 loop(State+1) % inc # of timeouts end. 44
- 45. Receiving messages Rubinius def actor_process() looping = true while looping Actor.receive do |鍖lter| 鍖lter.when Hello do |message| greet(message.from) end 鍖lter.when Goodbye do |message| message.from << KThxBye[self] looping = false end end end 45
- 46. Receiving messages Scala def act() loop { receive { case Hello(from) => greet(from) case Ping() => doPingThing() case Goodbye(from) => { from ! KThxBye(self) exit() } } } } 46
- 47. Receiving messages Io MyActor := Object clone MyActor hello := method(from, greet(from)) MyActor ping := method(doPingThing) MyActor goodbye := method(from, from @@kthxbye(self) ) anActor := MyActor clone // create actor instance anActor @@hello(self) anActor @@ping anActor @@goodbye(self) // look ma, no loops! 47
- 48. Actors avoid deadlock 48
- 49. No locks, no dead Agha(1985) says the semantics of the Actor model mean that deadlock doesnt exist in a syntactic sense Semantic deadlock can be detected and removed easily using wait-for graphs 49
- 50. Actors perform great! 50
- 51. Fine grained, hot CPU Lightweight context switching Actors dont suffer from locking granularity issues because there are no locks Actors can run full out on as many CPUs/machines as the underlying language can make available 51
- 52. Conclusion Light, fast processes needed to see huge scalability Erlangs the fastest and most mature (see OTP) You cant sneeze in the JVM without declaring its chemical composition Io is a beautiful little language 52
- 53. erlang is nice, because everything just works Paul Jenkins April 11, 2008 53
- 54. What next? Iolang... Io with: Support for multiple cores Easy way: one VM per core? Selective receive Support for remote actors 54
- 55. No cows were hurt... 55
- 56. Thanks for listening Picture credits Billie Joe: http://鍖ickr.com/photos/15034493@N00/112357008/sizes/l/ Joe Armstrong: http://鍖ickr.com/photos/mbiddulph/2037845171/sizes/l/ Evan Phoenix: http://鍖ickr.com/photos/scoop/1403258349/sizes/o/ Steve Dekorte: http://鍖ickr.com/photos/x180/276960722/sizes/o/ Martin Odersky: http://picasaweb.google.com/JavaPolis.com/JavaPolis2007/photo#5144254997958003122 Bruno: http://www.smh.com.au/ffximage/2006/10/31/bruno_wideweb__470x329,0.jpg Coady: http://鍖ickr.com/photos/sebastian_bergmann/116486823/sizes/o/ Ships: http://鍖ickr.com/photos/lesec/361956524/sizes/o/ Jim: http://鍖ickr.com/photos/kteague/278136366/sizes/o/ LOLcode: http://鍖ickr.com/photos/i-marco/534996407/sizes/o/ Triplets: http://鍖ickr.com/photos/origamijoel/217238954/sizes/l/ Spawn: http://鍖ickr.com/photos/austenhaines/399622840/ Parallel runners: http://鍖ickr.com/photos/moodeous/205711924/sizes/o/ Microscopic: http://鍖ickr.com/photos/braid44/2232461763/sizes/o/ Emoburger: http://鍖ickr.com/photos/gx9/269025682/ Cow: http://鍖ickr.com/photos/66164549@N00/542696674/sizes/o/ icanhascheezburger: http://icanhascheezburger.com/2007/01/11/i-can-has-cheezburger/ 56
- 57. Questions Make it quick, Pauls next 57
- 58. Dedicated to Cheryl, Cyan, Xavier, and Justin 58