Introduction
A proxy server is a computer program that acts as an intermediary between clients making requests to access resources and the servers that satisfy those requests by serving content. A web proxy is a special type of proxy server whose clients are typically web browsers and whose servers are the same servers that browsers use. When a web browser uses a proxy, it contacts the proxy instead of communicating directly with the web server; the proxy forwards its client’s request to the web server, reads the server’s response, then forwards the response to the client.
Proxies are useful for many purposes. Sometimes proxies are used in firewalls, so that browsers behind a firewall can only contact a server beyond the firewall via the proxy. A proxy may also perform translations on pages, for example, to make them viewable on web-enabled phones. Importantly, proxies are used as anonymizers: by stripping requests of all identifying information, a proxy can make the browser anonymous to web servers. Proxies can even be used to cache web objects by storing local copies of objects from servers then responding to future requests by reading them out of its cache rather than by communicating again with remote servers.
In this lab, you will write a simple HTTP proxy that caches web objects. For the first part of the lab, you will set up the proxy to accept incoming connections, read and parse requests, forward requests to web servers, read the servers’ responses, and forward those responses to the corresponding clients. This first part will involve learning about basic HTTP operation and how to use sockets to write programs that communicate over network connections. In the second part, you will upgrade your proxy to deal with multiple concurrent connections. This will introduce you to dealing with concurrency, a crucial systems concept. In the third and last part, you will add caching to your proxy using a simple main memory cache of recently accessed web content.
Logistics
This is an individual project. You may not use your grace days on this lab. You should do your work on one of the Shark machines.
Handout instructions
As usual, start by downloading the lab handout (proxylab-handout.tar) from Autolab and extracting it into the directory in which you plan to workissue the following command:
linux> tar xvf proxylab-handout.tar
If possible, extract the files directly onto a Shark machine; some operating systems and file transfer programs clobber Unix file system permission bits.
Robust I/O package
The handout directory contains the files csapp.c and csapp.h, which comprise the CS:APP package discussed in the CS:APP3e textbook. In addition to various error-handling wrapper functions and helper functions, the CS:APP package includes the robust I/O (RIO) package. When reading and writing socket data, you should use the RIO package instead of low-level I/O functions, such as read, write, or standard I/O functions, such as fread, and fwrite.
Keep in mind that the error-handling functions in CS:APP may not be appropriate for use in your proxy. Before blindly using wrapper functions or writing any of your own, carefully consider the proper action a server should take on each particular error.
Modularity
The skeleton file proxy.c is provided in the handout. proxy.c contains a main function that does practically nothing, and you should fill in that file with the guts of your implementation. Modularity, though, should be an important consideration, and it is permissible and encouraged for you to separate the individual modules of your implementation into different files. For example, your cache should be largely (or completely) decoupled from the rest of your proxy, so one popular idea is to move the implementation of the cache into separate files like cache.c and cache.h.
Makefile
Since you are free to add your own source files for this lab, you are responsible for updating the Makefile. The entire project should compile without warnings (you may want to use the -Werror flag), and you will want to determine the appropriate set of compilation flags (including optimization, linking, and debugging flags) for your final submission.
Part I: Implementing a sequential web proxy
The first step is implementing a basic sequential proxy that handles HTTP/1.0 GET requests. Other requests type, such as POST, are strictly optional.
When started, your proxy should listen for incoming connections on a port whose number will be specified on the command line. Once a connection is established, your proxy should read the entirety of the request from the client and parse the request. It should determine whether the client has sent a valid HTTP request; if so, it can then establish its own connection to the appropriate web server then request the object the client specified. Finally, your proxy should read the server’s response and forward it to the client.
HTTP/1.0 GET requests
When an end user enters a URL into the address bar of a web browser, the browser will send an HTTP request to the proxy that begins with a line.
In that case, the proxy should parse the request into at least the following fields: the hostname; and the path or query and everything following it, /hub/index.html. That way, the proxy can determine that it should open a connection and send an HTTP request of its own starting with a line of the following form:
GET /hub/index.html HTTP/1.0
Note that all lines in an HTTP request end with a carriage return, ‘\r’, followed by a newline, ‘\n’. Also important is that every HTTP request is terminated by an empty line: “\r\n”.
You should notice in the above example that the web browser’s request line ends with HTTP/1.1, while the proxy’s request line ends with HTTP/1.0. Modern web browsers will generate HTTP/1.1 requests, but your proxy should handle them and forward them as HTTP/1.0 requests.
It is important to consider that HTTP requests, even just the subset of HTTP/1.0 GET requests, can be incredibly complicated. The textbook describes certain details of HTTP transactions, but you should refer to RFC 1945 for the complete HTTP/1.0 specification. Ideally your HTTP request parser will be fully robust according to the relevant sections of RFC 1945, except for one detail: while the specification allows for multiline request fields, your proxy is not required to properly handle them. Of course, your proxy should never prematurely abort due to a malformed request.
Request headers
Request headers are very important elements of an HTTP request. Headers are essentially key-value pairs provided line-by-line following the first request line of an HTTP request. Of particular importance for this lab are the Host, User-Agent, Connection, and Proxy-Connection headers. Your proxy must perform the following operations with regard to the listed HTTP request headers:
- Always send a Host header. While this behavior is technically not sanctioned by the HTTP/1.0 specification, it is necessary to coax sensible responses out of certain web servers, especially those that use virtual hosting.
The Host header describes the hostname of the web server your proxy is trying to access. For example, to access your proxy would send the following header:
It is possible that web browsers will attach their own Host headers to their HTTP requests. If that is the case, your proxy should use the same Host header as the browser. - You may choose to always send the following User-Agent header
The header is provided on two separate lines because it does not fit as a single line in the writeup, but your proxy should send the header as a single line.
The User-Agent header identifies the client (in terms of parameters such as the operating system and browser), and web servers often use the identifying information to manipulate the content they serve. Sending this particular User-Agent: string may improve, in content and diversity, the material that you get back during simple telnet-style testing. - Always send the following Connection header:
Connection: close - Always send the following Proxy-Connection header:
Proxy-Connection: close
The Connection and Proxy-Connection headers are used to specify whether a connection will be kept alive after the first request/response exchange is completed. It is perfectly acceptable (and suggested) to have your proxy open a new connection for each request. Specifying close as the value of these headers alerts web servers that your proxy intends to close connections after the first request/response exchange.
With the exception of the Host header, your proxy should ignore the values of the request headers described above; instead, your proxy should always send the headers this document specifies.
For your convenience, the values of the described User-Agent header is provided to you as a string constant in proxy.c.
Finally, if a browser sends any additional request headers as part of an HTTP request, your proxy should forward them unchanged.
Port numbers
There are two significant classes of port numbers for this lab: HTTP request ports and your proxy’s listening port.
The HTTP request port is an optional field in the URL of an HTTP request. That is, the URL may be of the form, in which case your proxy should connect to the host on port 8080 instead of the default HTTP port, which is port 80. Your proxy must properly function whether or not the port number is included in the URL.
The listening port is the port on which your proxy should listen for incoming connections. Your proxy should accept a command line argument specifying the listening port number for your proxy. For example, with the following command, your proxy should listen for connections on port 12345: linux] ./proxy 12345
You will have to supply a port number every time you wish to test your proxy by running it. You may select any non-privileged port (greater than 1,024 and less than 65,536) as long as it is not used by other processes.
Since each proxy must use a unique listening port and many people will simultaneously be working on each machine, the script port-for-user.pl is provided to help you pick your own personal port number.
Use it to generate port number based on your Andrew ID:
linux> ./port-for-user.pl droh droh: 45806
The port, p, returned by port-for-user.pl is always an even number. So if you need an additional port number, say for the Tiny server, you can safely use ports p and p + 1.
Please don’t pick your own random port. If you do, you run the risk of interferring with another user.
Part II: Dealing with multiple concurrent requests
Production web proxies usually do not process requests sequentially; they process multiple requests in parallel. This is particularly important when handling a single request can involve a lengthy delay (as it might when contacting a remote web server). While your proxy waits for a response from the remote web server, it can work on a pending request from another client. Thus, once you have a working sequential proxy, you should alter it to simultaneously handle multiple requests.
POSIX Threads
You will use the POSIX Threads (Pthreads) library to spawn threads that will execute in parallel to serve multiple simultaneous requests. A simple way to implement concurrent request service is to spawn a new thread to process each new incoming request. In this architecture, the main server thread simply accepts connections and spawns off independent worker threads that deal with each request to completion and terminate when they are done. Other designs are also viable: you might alternatively decide to have your proxy create a pool of worker threads from the start. You may use any architecture you wish as long as your proxy exhibits true concurrency, but spawning a new worker thread for each request is the simplest and historically most common method.
The basic usage of Pthreads involves the implementation of a function that will serve as the start routine for new threads. Once a start routine exists, you can use the pthread_create function to create and start a new thread. New threads are by default joinable, which means that another thread must clean up spare resources after the thread exits, similar to how an exited process must be reaped by a call to wait.
Luckily, it is possible to detach threads, meaning spare resources are automatically reaped upon thread exit.
To properly detach threads, the first line of the start routine should be as follows:
pthread_detach(pthread_self());
Race conditions
While multithreading will almost certainly improve the performance of your web proxy, concurrency comes at a price: the threat of race conditions. Race conditions most often arise when there is a shared resource between multiple threads. You must find ways to avoid race conditions in your concurrent proxy. That will likely involve both minimizing shared resources and synchronizing access to shared resources. Synchronization involves the use of objects called locks, which come in many varieties. The Pthreads library contains all of the locking primitives you might need for synchronization in your proxy. See the your textbook for details.
Thread safety
The open clientfd and open listenfd functions described in the CS:APP3e textbook are thread safe are based on the modern and protocol-independent getaddrinfo function, and thus are thread safe.
Part III: Caching web objects
For the final part of the lab, you will add a cache to your proxy that will keep recently used Web objects in memory. HTTP actually defines a fairly complex model by which web servers can give instructions as to how the objects they serve should be cached and clients can specify how caches should be used on their behalf. However, your proxy will adopt a simplified approach.
When your proxy receives a web object from a server, it should cache it in memory as it transmits the object to the client. If another client requests the same object from the same server, your proxy need not reconnect to the server; it can simply resend the cached object.
Obviously, if your proxy were to cache every object that is ever requested, it would require an unlimited amount of memory. Moreover, because some web objects are larger than others, it might be the case that one giant object will consume the entire cache, preventing other objects from being cached at all. To avoid those problems, your proxy should have both a maximum cache size and a maximum cache object size.
Maximum cache size
The entirety of your proxy’s cache should have the following maximum size:
MAX_CACHE_SIZE = 1 MiB
When calculating the size of its cache, your proxy must only count bytes used to store the actual web objects; any extraneous bytes, including metadata, should be ignored.
Maximum object size
Your proxy should only cache web objects that do not exceed the following maximum size:
MAX_OBJECT_SIZE = 100 KiB
For your convenience, both size limits are provided as macros in proxy.c.
The easiest way to implement a correct cache is to allocate a buffer for each active connection and accumulate data as it is received from the server. If the size of the buffer ever exceeds the maximum object size, the buffer can be discarded. If the entirety of the web server’s response is read before the maximum object size is exceeded, then the object can be cached. Using this scheme, the maximum amount of data your proxy will ever use for web objects is the following, where T is the maximum number of active connections:
MAX_CACHE_SIZE + T * MAX_OBJECT_SIZE
Eviction policy
Your proxy’s cache should employ an eviction policy that approximates a least-recently-used (LRU) eviction policy. It doesn’t have to be strictly LRU, but it should be something reasonably close. Note that both reading an object and writing it count as using the object.
Synchronization
Accesses to the cache must be thread-safe, and ensuring that cache access is free of race conditions will likely be the more interesting aspect of this part of the lab. As a matter of fact, there is a special requirement that multiple threads must be able to simultaneously read from the cache. Of course, only one thread should be permitted to write to the cache at a time, but that restriction must not exist for readers.
As such, protecting accesses to the cache with one large exclusive lock is not an acceptable solution. You may want to explore options such as partitioning the cache, using Pthreads readers-writers locks, or using semaphores to implement your own readers-writers solution. In either, the fact that you don’t have to implement a strictly LRU eviction policy will give you some flexibility in supporting multiple readers.
Evaluation
This assignment will be graded out of a total of 100 points, which will be awarded based on the following criteria:
- BasicCorrectness: 40 points for basic proxy operation (autograded)
- Concurrency: 15 points for handling concurrent requests (autograded)
- Cache: 15 points for a working cache (autograded)
- RealPages: 20 points for correctly serving the following pages (5 pts each):
- Style: 10 points for style.
Autograding
Your handout materials include an autograder, called driver.sh, that Autolab will use to assign scores for BasicCorrectness, Concurrency, and Cache. From the proxylab-handout directory:
linux> ./driver.sh
You must run the driver on a Shark machine.
Manual testing
The autograder does only simple checks to confirm that your code is acting like a concurrent caching proxy.
Therefore, your TAs will do additional manual testing to see how your proxy deals with real pages of increasing complexity.
The TAs will test your proxy by running it on a Shark machine, and typing the RealPages URLs directly into the most recent version of the Firefox browser. We will not evaluate your proxy under any other conditions, or on any other pages.
Your TAs will also examine your code for any correctness issues that weren’t detected by the earlier tests. In particular, we will be looking for errors such as race conditions, thread safety issues, and non-approximate LRU cache implementations, memory and descriptor leaks, and improper SIGPIPE handling.
Robustness
As always, you must deliver a program that is robust to errors and even malformed or malicious input.
Servers are typically long-running processes, and web proxies are no exception. Think carefully about how long-running processes should react to different types of errors. For many kinds of errors, it is certainly inappropriate for your proxy to immediately exit.
Robustness implies other requirements as well, including invulnerability to error cases like segmentation faults and a lack of memory leaks and file descriptor leaks.
Style
Style points will be awarded based on the usual criteria. Proper error handling is as important as ever, and modularity is of particular importance for this lab, as there will be a significant amount of code. You should also strive for portability.
Testing and debugging
Besides the simple autograder, you will not have any sample inputs or a test program to test your implementation. You will have to come up with your own tests and perhaps even your own testing harness to help you debug your code and decide when you have a correct implementation. This is a valuable skill in the real world, where exact operating conditions are rarely known and reference solutions are often unavailable.
Fortunately there are many tools you can use to debug and test your proxy. Be sure to exercise all code paths and test a representative set of inputs, including base cases, typical cases, and edge cases.
thttpd
thttpd is a tiny HTTP server that is available for download on the web. You can run thttpd on any port given you have sufficient permissions, and it will serve files as requested. You can use thttpd to test your proxy’s ability to handle arbitrary content that you can make available as files to thttpd.
Tiny web server
Your handout directory the source code for the CS:APP Tiny web server. While not as powerful as thttpd, the CS:APP Tiny web server will be easy for you to modify as you see fit. It’s also a reasonable starting point for your proxy code. And it’s the server that the driver code uses to fetch pages.
curl
You can use curl to generate HTTP requests to any server, including your own proxy. It is an extremely useful debugging tool. For example, if your proxy and Tiny are both running on the local machine, Tiny is listening on port 15213, and proxy is listening on port 15214, then you can request a page from Tiny via your proxy.
telnet
As you saw during lecture, you can use telnet to open a connection to your proxy and send it HTTP requests.
netcat
netcat, also known as nc, is a versatile network utility. You can use netcat just like telnet, to open connections to servers. Hence, imagining that your proxy were running on catshark using port 12345 you can do something like the following to manually test your proxy
In addition to being able to connect to Web servers, netcat can also operate as a server itself. With the following command, you can run netcat as a server listening on port 12345.
Once you have set up a netcat server, you can generate a request to a phony object on it through your proxy, and you will be able to inspect the exact request that your proxy sent to netcat.
Web browsers
Eventually you should test your proxy using the most recent version of Mozilla Firefox. Visiting About Firefox will automatically update your browser to the most recent version.
It will be very exciting to see your proxy working through a real Web browser. Although the functionality of your proxy will be limited, you will notice that you are able to browse the vast majority of websites through your proxy.
An important caveat is that you must be very careful when testing caching using a Web browser. All modern
Web browsers have caches of their own, which you should disable before attempting to test your proxy’s cache.