Linux Network Programming, Part 1
The client code, shown in Listing 2, is a little simpler than the corresponding server code. To start the client, you must provide two command-line arguments: the host name or address of the machine the server is running on and the port number the server is bound to. Obviously, the server must be running before any client can connect to it.
In the client example (Listing 2), a socket is created like before. The first command-line argument is first assumed to be a host name for the purposes of finding the server's address. If this fails, it is then assumed to be a dotted-quad IP address. If this also fails, the client cannot resolve the server's address and will not be able to contact it.
Having located the server, an address structure is created for the client socket. No explicit call to bind() is needed here, as the connect() call handles all of this.
Once the connect() returns successfully, a duplex connection has been established. Like the server, the client can now use read() and write() calls to receive data on the connection.
Be aware of the following points when sending data over a socket connection:
Sending text is usually fine. Remember that different systems can have different conventions for the end of line (i.e., Unix is \012, whereas Microsoft uses \015\012).
Different architectures may use different byte-ordering for integers etc. Thankfully, the BSD guys thought of this problem already. There are routines (htons and nstoh for short integers, htonl and ntohl for long integers) which perform host-to-network order and network-to-host order conversions. Whether the network order is little-endian or big-endian doesn't really matter. It has been standardized across all TCP/IP network stack implementations. Unless you persistently pass only characters across sockets, you will run into byte-order problems if you do not use these routines. Depending on the machine architecture, these routines may be null macros or may actually be functional. Interestingly, a common source of bugs in socket programming is to forget to use these byte-ordering routines for filling the address field in the sock_addr structures. Perhaps it is not intuitively obvious, but this must also be done when using INADDR_ANY (i.e., htonl(INADDR_ANY)).
A key goal of network programming is to ensure processes do not interfere with each other in unexpected ways. In particular, servers must use appropriate mechanisms to serialize entry through critical sections of code, avoid deadlock and protect data validity.
You cannot (generally) pass a pointer to memory from one machine to another and expect to use it. It is unlikely you will want to do this.
Similarly, you cannot (generally) pass a file descriptor from one process to another (non-child) process via a socket and use it straightaway. Both BSD and SVR4 provide different ways of passing file descriptors between unrelated processes; however, the easiest way to do this in Linux is to use the /proc file system.
Additionally, you must ensure that you handle short writes correctly. Short writes happen when the write() call only partially writes a buffer to a file descriptor. They occur due to buffering in the operating system and to flow control in the underlying transport protocol. Certain system calls, termed slow system calls, may be interrupted. Some may or may not be automatically restarted, so you should explicitly handle this when network programming. The code excerpt in Listing 3 handles short writes.
Using multiple threads instead of multiple processes may lighten the load on the server host, thereby increasing efficiency. Context-switching between threads (in the same process address space) generally has much less associated overhead than switching between different processes. However, since most of the slave threads in this case are doing network I/O, they must be kernel-level threads. If they were user-level threads, the first thread to block on I/O would cause the whole process to block. This would result in starving all other threads of any CPU attention until the I/O had completed.
It is common to close unnecessary socket file descriptors in child and parent processes when using the simple forking model. This prevents the child or parent from potential erroneous reads or writes and also frees up descriptors, which are a limited resource. But do not try this when using threads. Multiple threads within a process share the same memory space and set of file descriptors. If you close the server socket in a slave thread, it closes for all other threads in that process.
Practical books for the most technical people on the planet. Newly available books include:
- Agile Product Development by Ted Schmidt
- Improve Business Processes with an Enterprise Job Scheduler by Mike Diehl
- Finding Your Way: Mapping Your Network to Improve Manageability by Bill Childers
- DIY Commerce Site by Reven Lerner
Plus many more.
- diff -u: What's New in Kernel Development
- Giving Silos Their Due
- What's New in 3D Printing, Part III: the Software
- Server Hardening
- 22 Years of Linux Journal on One DVD - Now Available
- Controversy at the Linux Foundation
- February 2016 Issue of Linux Journal
- Don't Burn Your Android Yet
- Firefox OS