Presentation is loading. Please wait.

Presentation is loading. Please wait.

Basic Networking & Interprocess Communication Vivek Pai Nov 27, 2001.

Similar presentations


Presentation on theme: "Basic Networking & Interprocess Communication Vivek Pai Nov 27, 2001."— Presentation transcript:

1 Basic Networking & Interprocess Communication Vivek Pai Nov 27, 2001

2 2 Everyone’s Getting Sick  Including me –Didn’t assign reading in syllabus –Did get rough grades computed –Still have one feedback missing –Putting off new project this week

3 3 Mechanics  Next project assigned next Tuesday  Only 5 projects instead of 6  Target: use threads, synchronization –Probably using webserver again –Not building on filesystem –Suggestions welcome  We’ve completely neglected the dining philosophers problem!

4 4 Dining Philosophers

5 5 Possible Solutions  Philosophers go in order  Place numbers on forks  If stuck, drop fork, try again  Defer by age or some other quantity  Stab each other

6 6 Deadlock “Solutions”  Eliminate parallelism  Order resources, grab in order  Determine priorities on contention  Restart & randomize Deadlock performance: cost of avoiding/detecting deadlock versus work frequency & work thrown away

7 7 Original Lecture Goals  Basic networking - Introduce the basics of networking, the new semantics versus standard file system calls, and how this affects the programming model. Discuss network basics such as naming, ports, connections, protocols, etc.  Interprocess communication - Show how “networking” is useful within a single machine to communicate data. Give examples of different domains, how they are implemented, and the effects within the kernel. Show how networking and interprocess communication can be used to allow easy distribution of applications.

8 8 Communication  You’ve already seen some of it –Web server project(s)  Machines have “names” –Human-readable names are convenience –“Actual” name is IP (Internet Protocol) address –For example, 127.0.0.1 means “this machine” –nslookup www.cs.princeton.edu gives 128.112.136.11www.cs.princeton.edu

9 9 Names & Services  Multiple protocols –ssh, ftp, telnet, http, etc. –How do we know how to connect?  Machines also have port numbers –16 bit quantity (0-65535) –Protocols have default port # –Can still do “telnet 128.112.136.11 80”

10 10 But The Web Is Massive  Possible names >> possible IP addresses –World population > possible IP addresses –Many names map to same IP addr –Use extra information to disambiguate –In HTTP, request contains “Host: name” header  Many connections to same (machine, port #) –Use (src addr, src port, dst addr, dst port) to identify connection

11 11 Circuit Switching versus Packet Switching  Circuit – reserve resources in advance –Hold resources for entire communication –Example: phone line  Packet – break data into small pieces –Pieces identify themselves, “share” links –Individual pieces routed to destination –Example: internet –Problem: no guarantee pieces reach

12 12 Who Got Rich By Packet Switching?

13 13 The “End To End” Argument  Don’t rely on lower layers of the system to ensure something happens  If it needs to occur, build the logic into the endpoints  Implications: –Intermediate components simplified –Repetition possible at endpoints (use OS)  What is reliability?

14 14 Do Applications Care?  Some do  Most don’t –Use whatever OS provides –Good enough for most purposes  What do applications want? –Performance –Simplicity

15 15 Reading & Writing  A file: –Is made into a “descriptor” via some call –Is an unstructured stream of bytes –Can be read/written –OS provides low-level interaction –Applications use read/write calls  Sounds workable?

16 16 Network Connections As FDs  Network connection usually called “socket”  Interesting new system calls –socket( ) – creates an fd for networking use –connect( ) – connects to the specified machine –bind( ) – specifies port # for this socket –listen( ) – waits for incoming connections –accept( ) – gets connection from other machine  And, of course, read( ) and write( )

17 17 New Semantics  Doing a write( ) –What’s the latency/bandwidth of a disk? –When does a write( ) “complete”? –Where did data actually go before? –Can we do something similar now?  What about read( ) –When should a read return? –What should a read return?

18 18 Buffering  Provided by OS –Memory on both sender and receiver sides –Sender: enables reliability, quick writes –Receiver: allows incoming data before read  Example – assume slow network –write(fd, buf, size); –memset(buf, 0, size) –write(fd, buf, size);

19 19 Interprocess Communications  Shared memory –Threads sharing address space –Processes memory-mapping the same file –Processes using shared memory system calls  Sockets and read/write –Nothing prohibits connection to same machine –Even faster mechanism – different “domain” –Unix domain (local) versus Internet (either)

20 20 Sockets vs Shared Memory  Sockets –Higher overhead –No common parent/file needed –Synchronous operation  Shared memory –Locking due to synchronous operation –Fast reads/writes – no OS intervention –Harder to use multiple machines

21 21 Even More Semantics  How do you express the following: –Do (task) until (message received) –Do (this task) until (receiver not ready) –Do (task) until (no more data)  Problem: implies knowing system behavior  Related: what happens when buffer fills/empties? (hint: think of filesystem)

22 22 Synchronous vs Asynchronous  Synchronous: do it now, wait until over  Asynchronous: start it now, check later Somewhat related:  Blocking: wait until it’s all done  Nonblocking: only do what can be done without blocking

23 23 Transferring Large Files  OS buffers are 16-64KB  Large files are >> buffer size  Assume two clients –Each requests a different large file –Both are on slow networks How do you design your server?

24 24 Server Design Choices  Processes –Each client handled by a different process  Threads –Each client handled by a different thread  Single process –Use nonblocking operations, multiplex

25 25 Processing Steps Read File Send Data Accept Conn Read Request Find File Send Header end

26 26 Blocking Steps Read File Send Data Accept Conn Read Request Find File Send Header end Network Blocking Disk Blocking

27 27 Concurrency Architecture  Overlap disk, network, & application-level processing  Architecture how steps are overlapped Note: implications for performance

28 28 Multiple Processes (MP) Pro: simple programming – rely on OS Cons:too many processes caching harder Process 1 Read File Send Data Accept Conn Read Request Find File Send Header Process N Read File Send Data Accept Conn Read Request Find File Send Header

29 29 Multiple Threads (MT) Pro: shared address space lower “context switch” overhead Cons:many threads requires kernel thread support synchronization needed Read File Send Data Accept Conn Read Request Find File Send Header

30 30 Single Process Event Driven (SPED) Pro: single address space no synchronization Cons:must explicitly handle pieces in practice, disk reads still block Read File Send Data Accept Conn Read Request Find File Send Header Event Dispatcher

31 31 Homework  Read about the select( ) system call  Reading from syllabus  I may send an e-mail with papers


Download ppt "Basic Networking & Interprocess Communication Vivek Pai Nov 27, 2001."

Similar presentations


Ads by Google