CSC364 Lecture 17

Final Review
(202605)

Transcript

1. Dr. Javier Gonzalez-Sanchez javiergs@calpoly.edu www.javiergs.info o ffi ce: 14 -227 CSC 364 Introduction to Networked, Distributed, and Parallel Computing Lecture 17. Final Review 1
2. 2 Grading Assignments L a bs + Quizzes Attend a nce & P a rticip a tion 20% 30% 10% 100% Project Fin a l Project Fin a l Ex a m 5% 20% Peer Ev a lu a tion 15%
3. 3 Grading >= 96.5 A >= 93 A- >= 89.5 B+ >= 86 B >= 82.5 B- >= 79 C+ >= 75.5 C >= 72 C- >= 68.5 D+ >= 65 D >= 61.5 D- < 61.5 F
4. 4 Disclaimer for Final Exam Review The f in a l ex a m is comprehensive, encomp a ssing a ll course m a teri a ls. This review covers only a subset of topics. Import a nt: Ensure you review a ll lecture slides a nd course m a teri a ls. The following slides include ex a mples rel a ted to selected topics. This is NOT a n exh a ustive list of ex a m topics.
5. Final Exam
6. 6 Key Topics Introduction to computing p a r a digms: • Networked Computing • Distributed Computing • P a r a llel Computing Systems sc a le beyond a single progr a m running on a single m a chine. Modern softw a re systems a re not just progr a ms but inter a cting components. Project-b a sed a ppro a ch, where concepts a re a pplied through progr a mming a ssignments.
7. Test Yourselves Networked Computing
8. 8 Client–Server Communication • Sockets provide a communic a tion ch a nnel • Ports • IP Address • DNS Address • loc a lhost • Client–server model: (1) Cre a te a socket, (2) Bind the socket to a port, (3) Listen for connections, (4) Accept a client connection, (5) Send/receive d a t a
9. 9 Client–Server Communication ServerSocket server = new ServerSocket(port); Socket client = server.accept(); InputStream in = client.getInputStream(); OutputStream out = client.getOutputStream();
10. 10 Client–Server Communication (5 points) Write a server program that sends the string 012345 to every client that connects. (5 points) Write a client program that connects to the server and receives the string 012345.
11. 11 Key Ideas • Protocols • TCP provides reli a ble communic a tion through byte stre a ms. • TCP Connections follow a lifecycle: connect → communic a te → close. • Systems m a y use short-lived or persistent connections. • Wh a t is HTTP? • HTTP request • HTTP response • URL - http://loc a lhost:8080/?n a me=J a vier&course=CSC364
12. HTTP
13. 13 HTTP Server ServerSocket server = new ServerSocket(8080); while(true) { Socket client = server.accept(); BufferedReader in = new BufferedReader( new InputStreamReader(client.getInputStream())); PrintWriter out = new PrintWriter(client.getOutputStream()); String request = in.readLine(); out.println("HTTP/1.1 200 OK"); out.println("Content-Type: text/plain"); out.println(); out.println("Hello from my Java Web Server”); client.close(); }
14. 14 HTTP Communication (10 points) Use an HTTP connection (Java code) to get the content of the page http://xyz.com
15. 15 Key Idea User types URL | v DNS lookup | v IP address | v TCP Socket Connection | v HTTP Request | v Server Application Logic | v HTTP Response | v Browser renders page
16. MQTT
17. 17 Key Ideas - how systems c a n decouple producers of inform a tion from consumers of inform a tion, a llowing components to communic a te without direct knowledge of e a ch other. - Broker (or mess a ge ch a nnel) - Topics
18. 18 MQTT MqttClient client = new MqttClient("tcp://localhost:1883", "publisher"); client.connect(); client.publish( "class/announcements", new MqttMessage("Exam tomorrow".getBytes())); MqttClient client = new MqttClient("tcp://localhost:1883", "subscriber"); client.connect(); client.subscribe( "class/announcements", (topic, msg) -> {println(new String(msg.getPayload()));} );
19. Test Yourselves Parallel Computing
20. 20 Key Ideas Wh a t is a Thre a d? Wh a t is a Process? Wh a t is a Progr a m?
21. 21 Runnable class ClientHandler implements Runnable { private Socket client; public ClientHandler(Socket client) { this.client = client; } public void run() { handleClient(client); } } new Thread(new ClientHandler(client)).start();
22. 22 Runnable (10 points) Implements the method handleClient. It creates a new thread (Worker) to take care of the new client. Implement the class Worker also. ServerSocket server = new ServerSocket(port); while (true) { Socket client = server.accept(); handleClient(client); }
23. 23 Key Ideas - Cre a ted a new thre a d for every client connection or h a ve a Pool - Thre a d explosion - Thre a d Pools - Java Executor Framework
24. 24 Thread Pool ExecutorService pool = Executors.newFixedThreadPool(10); pool.submit(task);
25. 25 Thread Pool (10 points) Re-implements the method handleClient (from a former question) using a thread pool: ServerSocket server = new ServerSocket(port); while (true) { Socket client = server.accept(); handleClient(client); }
26. 26 Locks lock.lock(); try { counter++; } finally { lock.unlock(); }
27. 27 Semaphore Semaphore semaphore = new Semaphore(3); semaphore.acquire(); try { // access resource } finally { semaphore.release(); }
28. 28 Key Ideas • Wh a t is a "r a ce condition”? • Wh a t is a “de a dlock”? • Wh a t is a “critic a l section”? • synchroniz a tion a nd (synchroniz a tion keyword in J a v a ) • locks vs. sem a phores • concurrent d a t a structures
29. Test Yourselves Distributed Computing
30. 30 Push and Pull Models Pull: Client ----request----> Server Client <---response---- Server Push: Producer ----event----> Subscriber Producer ----event----> Subscriber Producer ----event----> Subscriber
31. 31 Key Ideas • Concurrency design p a tterns provide structured solutions for coordin a ting thre a ds. • The Producer–Consumer p a ttern sep a r a tes d a t a gener a tion from processing. Producer → Bu ff er → Consumer • The Re a ders–Writers p a ttern m a n a ges concurrent a ccess to sh a red resources. Re a ders → sh a red resource ← Writers • Pipeline a rchitectures divide work into p a r a llel st a ges. Pipeline: St a ge 1 → St a ge 2 → St a ge 3 → St a ge 4 • Inmut a ble Objects
32. 32 Question(s) 1. Wh a t concurrency problem m a y occur? 2. Expl a in why the f in a l v a lue might not be 20000. 3. Rewrite increment() to m a ke it thre a d- s a fe using synchronized.
33. 33 Question(s) 1. Why is this implement a tion ine ff icient for m a ny re a ders? 2. Which concurrency p a ttern should be used here? 3. Implement it
34. 34 Question (10 points) You are designing a system where: • multiple sensors generate data continuously • several worker threads analyze the data • a monitoring dashboard must update whenever new results appear. Questions: 1. Which concurrency pattern should connect sensors and workers? 2. Which pattern should be used by the dashboard? 3. Draw a simple architecture diagram.
35. 35 Question H a ving this cl a ss:
36. 36 Question H a ving this cl a sses:
37. 37 Question H a ving this cl a sses:
38. 38 Question Question 1: Complete the run() method of the Re a der cl a ss so th a t it: • repe a tedly selects a r a ndom course code from codes • c a lls re a dCourse • prints the result • sleeps for 1000 ms Question 2: Using the s a me CourseC a t a log, complete the Writer cl a ss so th a t it: • upd a tes "CSC364" • writes "P a r a llel Computing v" + version • increments the version • sleeps for 3000 ms
39. 39 Question Question 3 (identify beh a vior) Suppose the progr a m cre a tes: • 5 re a der thre a ds • 2 writer thre a ds Answer the following: 1. C a n multiple re a ders run a t the s a me time? 2. C a n two writers upd a te the c a t a log a t the s a me time? 3. C a n a re a der a ccess the c a t a log while a writer holds the write lock?
40. 40 Top 20% Event-Driven Architecture (Blackboard) GUI in Java + Drawing Threads Listeners (for user input) Assignments and Final Project
41. 41 Questions
42. Lab.
43. 43 Let’s Work
44. CSC 364 Introduction to Introduction to Networked, Distributed, and Parallel Computing Javier Gonzalez-Sanchez, Ph.D. javiergs@calpoly.edu Winter 2026 Copyright. These slides can only be used as study material for the class CSC 364 at Cal Poly. They cannot be distributed or used for another purpose. 44