Take a look at the following function:
\nfunc getStatus(url string) (int, error) {\n resp, err := http.Get(url)\n if err != nil {\n return 0, err\n }\n defer resp.Body.Close()\n return resp.StatusCode, nil\n}It sends a GET request to url and returns the HTTP status code.
Using HTTP/1 (GODEBUG=http2client=0), calling this function 100 times against\nindex.golang.org takes ~710 milliseconds on Go 1.26\non my machine. On Go 1.27, the same program takes ~150 milliseconds.
That is almost 5 times faster! How is such a drastic performance difference possible?
\nTo understand the difference, consider what happens on the network when we make an HTTP request.
\nGo’s net/http package supports HTTP/1.0, HTTP/1.1, and\nHTTP/2. All three run over TCP, and usually over TLS when talking to HTTPS\nendpoints. Making a request over a new connection looks something like this:
TCP handshake (1 RTT) : the client sends aSYN packet, the server
replies withSYN-ACK , and the client replies withACK . Because the\nclient can send the first TLS message alongside that finalACK ,\nestablishing the TCP connection takes 1 round-trip time (RTT).
TLS handshake (2 RTT) : next, a TLS handshake exchanges encryption keys
and verifies the server’s identity; using TLS 1.2, this takes 2 RTT.
HTTP request : with the keys exchanged, HTTP requests can finally be sent
after a total of 3 RTT.
These round trips (plus some CPU time spent on TLS handshakes) add up quickly. In cross-region communications, one-way latencies often reach or exceed the 50 ms (100 ms RTT) shown in the diagram above. Compared to the 300 ms spent just setting up the connection, the HTTP request itself can be cheap to serve, taking only a little over 1 RTT when bandwidth is plentiful or the HTTP response is small.
\nEven in more favorable conditions, opening a new connection can still be\nexpensive. Our opening benchmark queries\nindex.golang.org, which is distributed worldwide and\nsupports TLS 1.3. This means that latency is relatively low, and the TLS\nhandshake can actually finish in just 1 RTT. Even so, the performance difference\nis still very noticeable.
\nTo avoid paying this setup cost on every request, HTTP clients and servers try to reuse existing connections across multiple requests when possible. The mechanics of this differ based on the HTTP version:
\nHTTP/1.0 : while not part of the official specification, clients and
servers often opt into connection reuse by sending aConnection: keep-alive header. Reuse continues until either endpoint omits the header\nor sends aConnection: close header.
HTTP/1.1 : connections are persistent by default, and can be reused for
subsequent requests and responses unless either side declares otherwise by\nsending aConnection: close header.
HTTP/2 : multiple requests and responses can be multiplexed over a single
TCP connection.
In HTTP/1 (both HTTP/1.0 and HTTP/1.1), multiple requests and responses on the\nsame connection are sent sequentially on the wire. If two GET requests are\nsent to index.golang.org over the same TCP\nconnection, the responses arrive as a single stream of bytes:
As shown in the diagram, net/http.Get reads and parses\nthe status line and headers from the byte stream to populate\nnet/http.Response, then returns immediately. The\nresponse body remains unread on the connection and is consumed only as the\ncaller reads from net/http.Response.Body.
Because responses are sent back-to-back on the same stream,\nnet/http can only find the start of the next response after\nthe current response body has been read to the very end (as determined by the\nContent-Length or Transfer-Encoding header).
Now look back at our example program, where we never read resp.Body:
func getStatus(url string) (int, error) {\n resp, err := http.Get(url)\n if err != nil {\n return 0, err\n }\n defer resp.Body.Close()\n return resp.StatusCode, nil\n}Through Go 1.26, when resp.Body.Close is called, net/http\nsees that there is still an unread response body. Realizing that it cannot reuse\nthe connection, it abandons the existing connection and creates new connections\nfor its next requests, causing it to repeatedly incur the cost of establishing\nnew connections.
(HTTP/2 does not have this problem. Because it multiplexes streams of frames over a single connection rather than sending responses strictly back-to-back, an unread response body on one stream does not block other requests from sharing the connection.)
\nThis subtlety of HTTP/1 connection reuse can cause silent performance degradation in seemingly innocuous code.
\nAs in our opening example, a client might only care about an API response’s\nstatus code. However, the API endpoint it is talking to might send bodies that\nprovide more detailed but ultimately unused information, such as {"success": true, "message": "update processed"}. This will unfortunately cause connection\nreuse to fail.
Even programs that do read the response body can run into this issue if they\nstop just short of the end of the body. For example,\nencoding/json.Decoder stops reading as soon as\nit sees the closing } of a JSON object. If a server sends a JSON response\nterminated with a newline, decoding the JSON object will leave the newline still\nunread. The connection cannot be reused for a new request until that newline is\nconsumed. Thus, a healthy connection might end up being abandoned, even though\nit was only a single byte away from being reusable.
The unread body problem has been a long-standing sharp edge. Over the years, many Go developers came up with their own workarounds. Most commonly, this is done by draining the response body prior to closing it:
\nfunc getStatus(url string) (int, error) {\n resp, err := http.Get(url)\n if err != nil {\n return 0, err\n }\n drainAndCloseBody(resp.Body)\n return resp.StatusCode, nil\n}\nfunc drainAndCloseBody(body io.ReadCloser) {\n io.Copy(io.Discard, body) // Drain the response body before closing it.\n body.Close()\n}While this works for small responses, if the unread body is large, establishing a new TCP connection can be faster than downloading and discarding the rest of the body. A better approach is to cap the number of bytes we are willing to drain:
\nfunc drainAndCloseBody(body io.ReadCloser) {\n io.CopyN(io.Discard, body, 256<<10) // Only drain up to 256 KB.\n body.Close()\n}However, a server might send a response body that trickles in slowly, so we also need to cap the time we spend draining:
\nfunc drainAndCloseBody(body io.ReadCloser) {\n drainDone := make(chan struct{}, 1)\n go func() {\n io.CopyN(io.Discard, body, 256<<10) // Only drain up to 256 KB.\n drainDone <- struct{}{}\n }()\n // Only spend at most 50 ms trying to drain the response body.\n select {\n case <-drainDone:\n case <-time.After(50 * time.Millisecond):\n // If we reach the 50 ms timeout, body.Close will run and interrupt the\n // io.CopyN, preventing goroutine leaks.\n }\n body.Close()\n}Of course, we also do not want to block for up to 50 milliseconds when closing a response body. Instead, we can drain the body asynchronously in the background and let it race against subsequent requests:
\nfunc drainAndCloseBody(body io.ReadCloser) {\n drainDone := make(chan struct{}, 1)\n go func() {\n io.CopyN(io.Discard, body, 256<<10) // Only drain up to 256 KB.\n drainDone <- struct{}{}\n }()\n go func() {\n // Only spend at most 50 ms trying to drain the response body.\n select {\n case <-drainDone:\n // If the drain can finish before a new request is sent, the\n // connection will be reused; otherwise, a new connection is\n // created, and the current connection will be put in a pool of\n // idle connections, potentially being reused for another future\n // request.\n case <-time.After(50 * time.Millisecond):\n // If we reach the 50 ms timeout, body.Close will run and interrupt\n // the io.CopyN, preventing goroutine leaks.\n }\n body.Close()\n }()\n}As we can see, draining response bodies safely is surprisingly tricky and cumbersome. Worse, as the practice of draining response bodies spread, some nuances were often lost. This has led to misunderstandings such as attempting to drain response bodies for HTTP/2 connections, when doing so is unnecessary.
\nStarting in Go 1.27, net/http.Transport\nautomatically does everything we have implemented in the code above: it will\nasynchronously drain an unread HTTP/1 response body for up to 50 ms or 256 KB.\nIt also implements additional optimizations; for example, when a response\ncontains the Connection: close header, draining the body is guaranteed to be\nwasted work, and it therefore skips doing so. Note that this behavior might\nevolve in future versions of Go, as other possible enhancements become apparent.
As a general best practice, we now recommend that users simply close their HTTP bodies once they no longer need them, like in our original example:
\nfunc getStatus(url string) (int, error) {\n resp, err := http.Get(url)\n if err != nil {\n return 0, err\n }\n defer resp.Body.Close()\n return resp.StatusCode, nil\n}Go is all about simplicity. Let your code stay clean and simple, and let\nnet/http deal with the complexities on your behalf.