HTTP pipelining

The pipelining of requests results in a dramatic improvement[4] in the loading times of HTML pages, especially over high latency connections such as satellite Internet connections. The speedup is less apparent on broadband connections, as the limitation of HTTP 1.1 still applies: the server must send its responses in the same order that the requests were received—so the entire connection remains first-in-first-out[1] and HOL blocking can occur.

The asynchronous operations of HTTP/2 and SPDY are solution for this.[5] By 2017 most browsers supported HTTP/2 by default which uses multiplexing instead.[2]

Non-idempotent requests such as POST should not be pipelined.[6] Read requests like GET and HEAD can always be pipelined. A sequence of other idempotent requests like PUT and DELETE can be pipelined or not depending on whether requests in the sequence depend on the effect of others.[1]

HTTP pipelining requires both the client and the server to support it. HTTP/1.1 conforming servers are required to produce valid responses to pipelined requests, but may not actually process requests concurrently.[7]

Most pipelining problems happen in HTTP intermediate nodes (hop-by-hop), i.e. in proxy servers, especially in transparent proxy servers (if one of them along the HTTP chain does not handle pipelined requests properly then nothing works as it should).[8]

Using pipelining with HTTP proxy servers is usually not recommended also because the HOL blocking problem may really slow down proxy server responses (as the server responses must be in the same order of the received requests).[1] [9]

Example: if a client sends 4 pipelined GET requests to a proxy through a single connection and the first one is not in its cache then the proxy has to forward that request to the destination web server; if the following three requests are instead found in its cache, the proxy has to wait for the web server response, then it has to send it to the client and only then it can send the three cached responses too.

If instead a client opens 4 connections to a proxy and sends 1 GET request per connection (without using pipelining) the proxy can send the three cached responses to client in parallel before the response from server is received, decreasing the overall completion time (because requests are served in parallel with no head-of-line blocking problem).[10] The same advantage exists in HTTP/2 multiplexed streams.

Pipelining was introduced in HTTP/1.1 and was not present in HTTP/1.0.[11]

There have always been complaints about browsers, proxy servers, etc. not working well when using pipelined requests / responses, up to the point that for many years (at least till 2011) software developers, engineers, web experts, etc. tried to summarize the various kind of problems they noted, to fix things and to give advices about how to deal with pipelining on the Open Web.[8]

• HTTP persistent connection
• WebSocket
• SPDY

• RFC 7230 Fielding, R.; Reschke, J. (2014). Fielding, R; Reschke, J (eds.). "Hypertext Transfer Protocol (HTTP/1.1): Message Syntax and Routing". ietf.org. doi:10.17487/RFC7230. Retrieved 2014-07-24. {{cite journal}}: Cite journal requires |journal= (help)
• HTTP/1.1 Pipelining FAQ at mozilla.org
• "Network Performance Effects of HTTP/1.1, CSS1, and PNG" at w3.org
• "Optimizing Page Load Times" article
• phttpget
• serf C library