The first two sections of this chapter provided a fire-hose tour of TCP connections and their performance implications. If you'd like to learn more about TCP networking, check out the resources listed at the end of the chapter.
We're going to switch gears now and get squarely back to HTTP. The rest of this chapter explains the HTTP technology for manipulating and optimizing connections. We'll start with the HTTP Connection header, an often misunderstood but important part of HTTP connection management. Then we'll talk about HTTP's connection optimization techniques.
HTTP allows a chain of HTTP intermediaries between the client and the ultimate origin server (proxies, caches, etc.). HTTP messages are forwarded hop by hop from the client, through intermediary devices, to the origin server (or the reverse).
In some cases, two adjacent HTTP applications may want to apply a set of options to their shared connection. The HTTP Connection header field has a comma-separated list of connection tokens that specify options for the connection that aren't propagated to other connections. For example, a connection that must be closed after sending the next message can be indicated by Connection: close.
The Connection header sometimes is confusing, because it can carry three different types of tokens:
HTTP header field names, listing headers relevant for only this connection
Arbitrary token values, describing nonstandard options for this connection
close, indicating the persistent
connection will be closed when done
If a connection token contains the name of an HTTP header field, that header field contains connection-specific information and must not be forwarded. Any header fields listed in the Connection header must be deleted before the message is forwarded. Placing a hop-by-hop header name in a Connection header is known as "protecting the header," because the Connection header protects against accidental forwarding of the local header. An example is shown in Figure 4-9.
When an HTTP application receives a message with a Connection header, the receiver parses and applies all options requested by the sender. It then deletes the Connection header and all headers listed in the Connection header before forwarding the message to the next hop. In addition, there are a few hop-by-hop headers that might not be listed as values of a Connection header, but must not be proxied. These include Proxy-Authenticate, Proxy-Connection, Transfer-Encoding, and Upgrade. For more about the Connection header, see Appendix C.
TCP performance delays can add up if the connections are managed naively. For example, suppose you have a web page with three embedded images. Your browser needs to issue four HTTP transactions to display this page: one for the top-level HTML and three for the embedded images. If each transaction requires a new connection, the connection and slow-start delays can add up (see Figure 4-10).
In addition to the real delay imposed by serial loading, there is also a psychological perception of slowness when a single image is loading and nothing is happening on the rest of the page. Users prefer multiple images to load at the same time.
Another disadvantage of serial loading is that some browsers are unable to display anything onscreen until enough objects are loaded, because they don't know the sizes of the objects until they are loaded, and they may need the size information to decide where to position the objects on the screen. In this situation, the browser may be making good progress loading objects serially, but the user may be faced with a blank white screen, unaware that any progress is being made at all.
Several current and emerging techniques are available to improve HTTP connection performance. The next several sections discuss four such techniques:
Concurrent HTTP requests across multiple TCP connections
Reusing TCP connections to eliminate connect/close delays
Concurrent HTTP requests across a shared TCP connection
Interleaving chunks of requests and responses (experimental)
 For the purpose of this example, assume all objects are roughly the same size and are hosted from the same server, and that the DNS entry is cached, eliminating the DNS lookup time.
 This is true even if loading multiple images at the same time is slower than loading images one at a time! Users often perceive multiple-image loading as faster.
 HTML designers can help eliminate this "layout delay" by explicitly adding width and height attributes to HTML tags for embedded objects such as images. Explicitly providing the width and height of the embedded image allows the browser to make graphical layout decisions before it receives the objects from the server.