0
Fork 0
mirror of https://github.com/ninenines/cowboy.git synced 2025-07-14 04:10:24 +00:00
Small, fast, modern HTTP server for Erlang/OTP.
Find a file
Loïc Hoguin 49be0f57cf
Implement dynamic socket buffer sizes
Cowboy will set the socket's buffer size dynamically to
better fit the current workload. When the incoming data
is small, a low buffer size reduces the memory footprint
and improves responsiveness and therefore performance.
When the incoming data is large, such as large HTTP
request bodies, a larger buffer size helps us avoid
doing too many binary appends and related allocations.

Setting a large buffer size for all use cases is
sub-optimal because allocating more than needed
necessarily results in a performance hit (not just
increased memory usage).

By default Cowboy starts with a buffer size of 8192 bytes.
It then doubles or halves the buffer size depending on
the size of the data it receives from the socket. It
stops decreasing at 8192 and increasing at 131072 by
default.

To keep track of the size of the incoming data Cowboy
maintains a moving average. It allows Cowboy to avoid
changing the buffer too often but still react quickly
when necessary. Cowboy will increase the buffer size
when the moving average is above 90% of the current
buffer size, and decrease when the moving average is
below 40% of the current buffer size.

The current buffer size and moving average are
propagated when switching protocols. The dynamic buffer
is implemented in HTTP/1, HTTP/2 and HTTP/1 Websocket.
HTTP/2 Websocket has it disabled because it doesn't
interact directly with the socket; in that case it
is HTTP/2 that has a dynamic buffer.

The dynamic buffer provides a very large performance improvement
in many scenarios, at minimal cost for others. Because it largely
depend on the underlying protocol the improvements are no all equal.
TLS and compression also impact the results.

The improvement when reading a large request body, with the
requests repeated in a fast loop are:

* HTTP: 6x to 20x faster
* HTTPS: 2x to 6x faster
* H2: 4x to 5x faster
* H2C: 20x to 40x faster

I am not sure why H2C's performance was so bad, especially compared
to H2, when using default buffer sizes. Dynamic buffers make H2C a
lot more viable with default settings.

The performance impact on "hello world" type requests is minimal,
it goes from -5% to +5% roughly.

Websocket improvements vary again depending on the protocol, but
also depending on whether compression is enabled:

* HTTP echo: roughly 2x faster
* HTTP send: roughly 4x faster
* H2C echo: roughly 2x faster
* H2C send: 3x to 4x faster

In the echo test we reply back, and Gun doesn't have the dynamic
buffer optimisation, so that probably explains the x2 difference.

With compression however there isn't much improvement. The results
are roughly within -10% to +10% of each other. Zlib compression
seems to be a bottleneck, or at least to modify the performance
profile to such an extent that the size of the buffer does not
matter. This happens to randomly generated binary data as well
so it is probably not caused by the test data.
2025-02-05 14:29:58 +01:00
.github/workflows CI: Remove unneeded env 2024-11-08 12:54:26 +01:00
doc/src Implement dynamic socket buffer sizes 2025-02-05 14:29:58 +01:00
ebin Initial HTTP/3 implementation 2024-03-26 15:53:48 +01:00
examples Fix markdown example for OTP-27 2024-11-07 13:28:26 +01:00
src Implement dynamic socket buffer sizes 2025-02-05 14:29:58 +01:00
test Implement dynamic socket buffer sizes 2025-02-05 14:29:58 +01:00
.gitattributes Add ws_perf_SUITE to measure Websocket performance 2025-01-15 13:28:57 +01:00
.gitignore gitignore: add examples/ dependency files 2016-05-25 12:54:12 +02:00
CONTRIBUTING.asciidoc Update CONTRIBUTING.asciidoc 2024-11-15 11:16:45 +01:00
erlang.mk Update Erlang.mk 2024-11-06 16:33:39 +01:00
LICENSE Refresh copyright lines 2024-01-25 11:22:54 +01:00
Makefile Avoid resetting Websocket idle_timeout timer too often 2025-01-15 13:31:19 +01:00
plugins.mk Document the commands based Websocket interface 2019-10-06 16:51:27 +02:00
README.asciidoc Update Discord link to a permanent invite 2024-11-27 12:10:12 +01:00
rebar.config Add WS compression test where only server sets client_max_window_bits 2025-01-23 11:39:48 +01:00

= Cowboy

Cowboy is a small, fast and modern HTTP server for Erlang/OTP.

== Goals

Cowboy aims to provide a *complete* HTTP stack in a *small* code base.
It is optimized for *low latency* and *low memory usage*, in part
because it uses *binary strings*.

Cowboy provides *routing* capabilities, selectively dispatching requests
to handlers written in Erlang.

Because it uses Ranch for managing connections, Cowboy can easily be
*embedded* in any other application.

Cowboy is *clean* and *well tested* Erlang code.

== Online documentation

* https://ninenines.eu/docs/en/cowboy/2.12/guide[User guide]
* https://ninenines.eu/docs/en/cowboy/2.12/manual[Function reference]

== Offline documentation

* While still online, run `make docs`
* User guide available in `doc/` in PDF and HTML formats
* Function reference man pages available in `doc/man3/` and `doc/man7/`
* Run `make install-docs` to install man pages on your system
* Full documentation in Asciidoc available in `doc/src/`
* Examples available in `examples/`

== Getting help

* https://discord.gg/x25nNq2fFE[Discord server]
* https://github.com/ninenines/cowboy/issues[Issues tracker]
* https://ninenines.eu/services[Commercial Support]
* https://github.com/sponsors/essen[Sponsor me!]