English | 🇨🇳中文
gnet
is an event-driven networking framework that is fast and small. It makes direct epoll and kqueue syscalls rather than using the standard Go net package, and works in a similar manner as netty and libuv.
The goal of this project is to create a server framework for Go that performs on par with Redis and Haproxy for packet handling.
gnet
sells itself as a high-performance, lightweight, non-blocking, event-driven networking framework written in pure Go which works on transport layer with TCP/UDP/Unix-Socket protocols, so it allows developers to implement their own protocols(HTTP, RPC, WebSocket, Redis, etc.) of application layer upon gnet
for building diversified network applications, for instance, you get an HTTP Server or Web Framework if you implement HTTP protocol upon gnet
while you have a Redis Server done with the implementation of Redis protocol upon gnet
and so on.
gnet
derives from the project: evio
while having a much higher performance.
- High-performance event-loop under networking model of multiple threads/goroutines
- Built-in load balancing algorithm: Round-Robin
- Built-in goroutine pool powered by the library ants
- Built-in memory pool with bytes powered by the library pool
- Concise APIs
- Efficient memory usage: Ring-Buffer
- Supporting multiple protocols: TCP, UDP, and Unix Sockets
- Supporting two event-driven mechanisms: epoll on Linux and kqueue on FreeBSD
- Supporting asynchronous write operation
- Flexible ticker event
- SO_REUSEPORT socket option
- Built-in multiple codecs to encode/decode network frames into/from TCP stream: LineBasedFrameCodec, DelimiterBasedFrameCodec, FixedLengthFrameCodec and LengthFieldBasedFrameCodec, referencing netty codec, also supporting customized codecs
- Supporting Windows platform with
event-driven mechanism of IOCPGo stdlib: net - Additional load-balancing algorithms: Random, Least-Connections, Consistent-hashing and so on
- TLS support
- Implementation of
gnet
Client
gnet
redesigns and implements a new built-in networking model of multiple threads/goroutines: 『multiple reactors』 which is also the default networking model of multiple threads of netty
, Here's the schematic diagram:
and it works as the following sequence diagram:
You may ask me a question: what if my business logic in EventHandler.React
contains some blocking code which leads to blocking in event-loop of gnet
, what is the solution for this kind of situation?
As you know, there is a most important tenet when writing code under gnet
: you should never block the event-loop in the EventHandler.React
, otherwise, it will lead to a low throughput in your gnet
server, which is also the most important tenet in netty
.
And the solution to that could be found in the subsequent networking model of multiple threads/goroutines of gnet
: 『multiple reactors with thread/goroutine pool』which pulls you out from the blocking mire, it will construct a worker-pool with fixed capacity and put those blocking jobs in EventHandler.React
into the worker-pool to make the event-loop goroutines non-blocking.
The architecture diagram of networking model:『multiple reactors with thread/goroutine pool』architecture is in here:
and it works as the following sequence diagram:
gnet
implements the networking model:『multiple reactors with thread/goroutine pool』by the aid of a high-performance goroutine pool called ants that allows you to manage and recycle a massive number of goroutines in your concurrent programs, the full features and usages in ants
are documented here.
gnet
integrates ants
and provides the pool.NewWorkerPool
method that you can invoke to instantiate a ants
pool where you are able to put your blocking code logic in EventHandler.React
and invoke the function of gnet.Conn.AsyncWrite
to send out data asynchronously in worker pool after you finish the blocking process and get the output data, which makes the goroutine of event-loop non-blocking.
The details about integrating gnet
with ants
are shown here.
There are two ring-buffers inside gnet
: inbound buffer and outbound buffer to buffer and manage inbound/outbound network data.
The purpose of implementing inbound and outbound ring-buffers in gnet
is to transfer the logic of buffering and managing network data based on application protocol upon TCP stream from business server to framework and unify the network data buffer, which minimizes the complexity of business code so that developers are able to concentrate on business logic instead of the underlying function.
gnet
requires Go 1.9 or later.
go get -u github.com/panjf2000/gnet
gnet
is available as a Go module, with Go 1.11 Modules support (Go 1.11+), just simply import "github.com/panjf2000/gnet"
in your source code and go [build|run|test]
will download the necessary dependencies automatically.
The detailed documentation is located in here: docs of gnet, but let's pass through the brief instructions first.
It is easy to create a network server with gnet
. All you have to do is just to make your implementation of gnet.EventHandler
interface and register your event-handler functions to it, then pass it to the gnet.Serve
function along with the binding address(es). Each connection is represented as a gnet.Conn
interface that is passed to various events to differentiate the clients. At any point you can close a client or shutdown the server by return a Close
or Shutdown
action from an event.
The simplest example to get you started playing with gnet
would be the echo server. So here you are, a simplest echo server upon gnet
that is listening on port 9000:
package main
import (
"log"
"github.com/panjf2000/gnet"
)
type echoServer struct {
*gnet.EventServer
}
func (es *echoServer) React(c gnet.Conn) (out []byte, action gnet.Action) {
out = c.Read()
c.ResetBuffer()
return
}
func main() {
echo := new(echoServer)
log.Fatal(gnet.Serve(echo, "tcp://:9000", gnet.WithMulticore(true)))
}
As you can see, this example of echo server only sets up the EventHandler.React
function where you commonly write your main business code and it will be invoked once the server receives input data from a client. The output data will be then sent back to that client by assigning the out
variable and return it after your business code finish processing data(in this case, it just echo the data back).
package main
import (
"log"
"time"
"github.com/panjf2000/gnet"
"github.com/panjf2000/gnet/pool"
)
type echoServer struct {
*gnet.EventServer
pool *pool.WorkerPool
}
func (es *echoServer) React(c gnet.Conn) (out []byte, action gnet.Action) {
data := append([]byte{}, c.Read()...)
c.ResetBuffer()
// Use ants pool to unblock the event-loop.
_ = es.pool.Submit(func() {
time.Sleep(1 * time.Second)
c.AsyncWrite(data)
})
return
}
func main() {
p := pool.NewWorkerPool()
defer p.Release()
echo := &echoServer{pool: p}
log.Fatal(gnet.Serve(echo, "tcp://:9000", gnet.WithMulticore(true)))
}
Like I said in the 『Multiple Reactors + Goroutine-Pool』section, if there are blocking code in your business logic, then you ought to turn them into non-blocking code in any way, for instance you can wrap them into a goroutine, but it will result in a massive amount of goroutines if massive traffic is passing through your server so I would suggest you utilize a goroutine pool like ants to manage those goroutines and reduce the cost of system resources.
All gnet examples:
Echo Server
package main
import (
"flag"
"fmt"
"log"
"github.com/panjf2000/gnet"
)
type echoServer struct {
*gnet.EventServer
}
func (es *echoServer) OnInitComplete(srv gnet.Server) (action gnet.Action) {
log.Printf("Echo server is listening on %s (multi-cores: %t, loops: %d)\n",
srv.Addr.String(), srv.Multicore, srv.NumLoops)
return
}
func (es *echoServer) React(c gnet.Conn) (out []byte, action gnet.Action) {
out = c.Read()
c.ResetBuffer()
return
}
func main() {
var port int
var multicore bool
// Example command: go run echo.go --port 9000 --multicore true
flag.IntVar(&port, "port", 9000, "server port")
flag.BoolVar(&multicore, "multicore", true, "multicore")
flag.Parse()
echo := new(echoServer)
log.Fatal(gnet.Serve(echo, fmt.Sprintf("tcp://:%d", port), gnet.WithMulticore(multicore)))
}
HTTP Server
package main
import (
"bytes"
"flag"
"fmt"
"log"
"os"
"strconv"
"strings"
"time"
"github.com/panjf2000/gnet"
)
var res string
type request struct {
proto, method string
path, query string
head, body string
remoteAddr string
}
type httpServer struct {
*gnet.EventServer
noparse bool
}
func (hs *httpServer) OnInitComplete(srv gnet.Server) (action gnet.Action) {
log.Printf("HTTP server is listening on %s (multi-cores: %t, loops: %d)\n",
srv.Addr.String(), srv.Multicore, srv.NumLoops)
return
}
func (hs *httpServer) React(c gnet.Conn) (out []byte, action gnet.Action) {
data := c.Read()
if hs.noparse && bytes.Contains(data, []byte("\r\n\r\n")) {
// for testing minimal single packet request -> response.
out = appendresp(nil, "200 OK", "", res)
c.ResetBuffer()
return
}
// process the pipeline
var req request
leftover, err := parsereq(data, &req)
if err != nil {
// bad thing happened
out = appendresp(out, "500 Error", "", err.Error()+"\n")
action = gnet.Close
return
} else if len(leftover) == len(data) {
// request not ready, yet
return
}
// handle the request
req.remoteAddr = c.RemoteAddr().String()
out = appendhandle(out, &req)
c.ResetBuffer()
return
}
func main() {
var port int
var multicore bool
var aaaa bool
var noparse bool
// Example command: go run http.go --port 8080 --multicore true
flag.IntVar(&port, "port", 8080, "server port")
flag.BoolVar(&aaaa, "aaaa", false, "aaaaa....")
flag.BoolVar(&noparse, "noparse", true, "do not parse requests")
flag.BoolVar(&multicore, "multicore", true, "multicore")
flag.Parse()
if os.Getenv("NOPARSE") == "1" {
noparse = true
}
if aaaa {
res = strings.Repeat("a", 1024)
} else {
res = "Hello World!\r\n"
}
http := &httpServer{noparse: noparse}
// We at least want the single http address.
addr := fmt.Sprintf("tcp"+"://:%d", port)
// Start serving!
log.Fatal(gnet.Serve(http, addr, gnet.WithMulticore(multicore)))
}
// appendhandle handles the incoming request and appends the response to
// the provided bytes, which is then returned to the caller.
func appendhandle(b []byte, req *request) []byte {
return appendresp(b, "200 OK", "", res)
}
// appendresp will append a valid http response to the provide bytes.
// The status param should be the code plus text such as "200 OK".
// The head parameter should be a series of lines ending with "\r\n" or empty.
func appendresp(b []byte, status, head, body string) []byte {
b = append(b, "HTTP/1.1"...)
b = append(b, ' ')
b = append(b, status...)
b = append(b, '\r', '\n')
b = append(b, "Server: gnet\r\n"...)
b = append(b, "Date: "...)
b = time.Now().AppendFormat(b, "Mon, 02 Jan 2006 15:04:05 GMT")
b = append(b, '\r', '\n')
if len(body) > 0 {
b = append(b, "Content-Length: "...)
b = strconv.AppendInt(b, int64(len(body)), 10)
b = append(b, '\r', '\n')
}
b = append(b, head...)
b = append(b, '\r', '\n')
if len(body) > 0 {
b = append(b, body...)
}
return b
}
// parsereq is a very simple http request parser. This operation
// waits for the entire payload to be buffered before returning a
// valid request.
func parsereq(data []byte, req *request) (leftover []byte, err error) {
sdata := string(data)
var i, s int
var head string
var clen int
var q = -1
// method, path, proto line
for ; i < len(sdata); i++ {
if sdata[i] == ' ' {
req.method = sdata[s:i]
for i, s = i+1, i+1; i < len(sdata); i++ {
if sdata[i] == '?' && q == -1 {
q = i - s
} else if sdata[i] == ' ' {
if q != -1 {
req.path = sdata[s:q]
req.query = req.path[q+1 : i]
} else {
req.path = sdata[s:i]
}
for i, s = i+1, i+1; i < len(sdata); i++ {
if sdata[i] == '\n' && sdata[i-1] == '\r' {
req.proto = sdata[s:i]
i, s = i+1, i+1
break
}
}
break
}
}
break
}
}
if req.proto == "" {
return data, fmt.Errorf("malformed request")
}
head = sdata[:s]
for ; i < len(sdata); i++ {
if i > 1 && sdata[i] == '\n' && sdata[i-1] == '\r' {
line := sdata[s : i-1]
s = i + 1
if line == "" {
req.head = sdata[len(head)+2 : i+1]
i++
if clen > 0 {
if len(sdata[i:]) < clen {
break
}
req.body = sdata[i : i+clen]
i += clen
}
return data[i:], nil
}
if strings.HasPrefix(line, "Content-Length:") {
n, err := strconv.ParseInt(strings.TrimSpace(line[len("Content-Length:"):]), 10, 64)
if err == nil {
clen = int(n)
}
}
}
}
// not enough data
return data, nil
}
Push Server
package main
import (
"flag"
"fmt"
"log"
"sync"
"time"
"github.com/panjf2000/gnet"
)
type pushServer struct {
*gnet.EventServer
tick time.Duration
connectedSockets sync.Map
}
func (ps *pushServer) OnInitComplete(srv gnet.Server) (action gnet.Action) {
log.Printf("Push server is listening on %s (multi-cores: %t, loops: %d), "+
"pushing data every %s ...\n", srv.Addr.String(), srv.Multicore, srv.NumLoops, ps.tick.String())
return
}
func (ps *pushServer) OnOpened(c gnet.Conn) (out []byte, action gnet.Action) {
log.Printf("Socket with addr: %s has been opened...\n", c.RemoteAddr().String())
ps.connectedSockets.Store(c.RemoteAddr().String(), c)
return
}
func (ps *pushServer) OnClosed(c gnet.Conn, err error) (action gnet.Action) {
log.Printf("Socket with addr: %s is closing...\n", c.RemoteAddr().String())
ps.connectedSockets.Delete(c.RemoteAddr().String())
return
}
func (ps *pushServer) Tick() (delay time.Duration, action gnet.Action) {
log.Println("It's time to push data to clients!!!")
ps.connectedSockets.Range(func(key, value interface{}) bool {
addr := key.(string)
c := value.(gnet.Conn)
c.AsyncWrite([]byte(fmt.Sprintf("heart beating to %s\n", addr)))
return true
})
delay = ps.tick
return
}
func (ps *pushServer) React(c gnet.Conn) (out []byte, action gnet.Action) {
out = c.Read()
c.ResetBuffer()
return
}
func main() {
var port int
var multicore bool
var interval time.Duration
var ticker bool
// Example command: go run push.go --port 9000 --tick 1s
flag.IntVar(&port, "port", 9000, "server port")
flag.BoolVar(&multicore, "multicore", true, "multicore")
flag.DurationVar(&interval, "tick", 0, "pushing tick")
flag.Parse()
if interval > 0 {
ticker = true
}
push := &pushServer{tick: interval}
log.Fatal(gnet.Serve(push, fmt.Sprintf("tcp://:%d", port), gnet.WithMulticore(multicore), gnet.WithTicker(ticker)))
}
Codec Client/Server
Client:
// Reference https://github.com/smallnest/goframe/blob/master/_examples/goclient/client.go
package main
import (
"encoding/binary"
"fmt"
"net"
"github.com/smallnest/goframe"
)
func main() {
conn, err := net.Dial("tcp", "127.0.0.1:9000")
if err != nil {
panic(err)
}
defer conn.Close()
encoderConfig := goframe.EncoderConfig{
ByteOrder: binary.BigEndian,
LengthFieldLength: 4,
LengthAdjustment: 0,
LengthIncludesLengthFieldLength: false,
}
decoderConfig := goframe.DecoderConfig{
ByteOrder: binary.BigEndian,
LengthFieldOffset: 0,
LengthFieldLength: 4,
LengthAdjustment: 0,
InitialBytesToStrip: 4,
}
fc := goframe.NewLengthFieldBasedFrameConn(encoderConfig, decoderConfig, conn)
err = fc.WriteFrame([]byte("hello"))
if err != nil {
panic(err)
}
err = fc.WriteFrame([]byte("world"))
if err != nil {
panic(err)
}
buf, err := fc.ReadFrame()
if err != nil {
panic(err)
}
fmt.Println("received: ", string(buf))
buf, err = fc.ReadFrame()
if err != nil {
panic(err)
}
fmt.Println("received: ", string(buf))
}
Server:
package main
import (
"encoding/binary"
"flag"
"fmt"
"log"
"time"
"github.com/panjf2000/gnet"
"github.com/panjf2000/gnet/pool"
)
type codecServer struct {
*gnet.EventServer
addr string
multicore bool
async bool
codec gnet.ICodec
workerPool *pool.WorkerPool
}
func (cs *codecServer) OnInitComplete(srv gnet.Server) (action gnet.Action) {
log.Printf("Test codec server is listening on %s (multi-cores: %t, loops: %d)\n",
srv.Addr.String(), srv.Multicore, srv.NumLoops)
return
}
func (cs *codecServer) React(c gnet.Conn) (out []byte, action gnet.Action) {
if cs.async {
data := append([]byte{}, c.ReadFrame()...)
_ = cs.workerPool.Submit(func() {
c.AsyncWrite(data)
})
return
}
out = c.ReadFrame()
return
}
func testCodecServe(addr string, multicore, async bool, codec gnet.ICodec) {
var err error
if codec == nil {
encoderConfig := gnet.EncoderConfig{
ByteOrder: binary.BigEndian,
LengthFieldLength: 4,
LengthAdjustment: 0,
LengthIncludesLengthFieldLength: false,
}
decoderConfig := gnet.DecoderConfig{
ByteOrder: binary.BigEndian,
LengthFieldOffset: 0,
LengthFieldLength: 4,
LengthAdjustment: 0,
InitialBytesToStrip: 4,
}
codec = gnet.NewLengthFieldBasedFrameCodec(encoderConfig, decoderConfig)
}
cs := &codecServer{addr: addr, multicore: multicore, async: async, codec: codec, workerPool: pool.NewWorkerPool()}
err = gnet.Serve(cs, addr, gnet.WithMulticore(multicore), gnet.WithTCPKeepAlive(time.Minute*5), gnet.WithCodec(codec))
if err != nil {
panic(err)
}
}
func main() {
var port int
var multicore bool
// Example command: go run server.go --port 9000 --multicore true
flag.IntVar(&port, "port", 9000, "server port")
flag.BoolVar(&multicore, "multicore", true, "multicore")
flag.Parse()
addr := fmt.Sprintf("tcp://:%d", port)
testCodecServe(addr, true, false, nil)
}
For more details, check out here: examples of gnet.
Current supported I/O events in gnet
:
EventHandler.OnInitComplete
fires when the server has been initialized and ready to accept new connections.EventHandler.OnOpened
fires once a connection has been opened.EventHandler.OnClosed
fires after a connection has been closed.EventHandler.React
fires when the server receives inbound data from a socket/connection. (usually it is where you write the code of business logic)EventHandler.Tick
fires right after the server starts and then fires every specified interval.EventHandler.PreWrite
fires just before any data has been written to client.
The EventHandler.Tick
event fires ticks at a specified interval.
The first tick fires right after the gnet server starts up and if you intend to set up a ticker event, don't forget to pass an option: gnet.WithTicker(true)
to gnet.Serve
.
events.Tick = func() (delay time.Duration, action Action){
log.Printf("tick")
delay = time.Second
return
}
The gnet.Serve
function can bind to UDP addresses.
- All incoming and outgoing packets will not be buffered but sent individually.
- The
EventHandler.OnOpened
andEventHandler.OnClosed
events are not available for UDP sockets, only theReact
event.
The gnet.WithMulticore(true)
indicates whether the server will be effectively created with multi-cores, if so, then you must take care of synchronizing memory between all event callbacks, otherwise, it will run the server with a single thread. The number of threads in the server will be automatically assigned to the value of runtime.NumCPU()
.
The current built-in load balancing algorithm in gnet
is Round-Robin.
gnet
server is able to utilize the SO_REUSEPORT option which allows multiple sockets on the same host to bind to the same port and the OS kernel takes care of the load balancing for you, it wakes one socket per accpet
event coming to resolved the thundering herd
.
By default, gnet
is not going to be haunted by the thundering herd
under its networking model:『multiple reactors』which gets only one main reactor to listen on "address:port" and accept new sockets. So this SO_REUSEPORT
option is trivial in gnet
but note that it will fall back to the old networking model of evio
when you enable the SO_REUSEPORT
option.
Just use functional options to set up SO_REUSEPORT
and you can enjoy this feature:
gnet.Serve(events, "tcp://:9000", gnet.WithMulticore(true), gnet.WithReusePort(true)))
There are multiple built-in codecs in gnet
which allow you to encode/decode frames into/from TCP stream.
So far gnet
has four kinds of built-in codecs: LineBasedFrameCodec, DelimiterBasedFrameCodec, FixedLengthFrameCodec and LengthFieldBasedFrameCodec, which generally meets most scenarios, but still gnet
allows users to customize their own codecs in their gnet
servers by implementing the interface gnet.ICodec and replacing the default codec in gnet
with customized codec via functional options.
Here is an example with codec, showing you how to leverage codec to encode/decode network frames into/from TCP stream.
# Machine information
OS : Ubuntu 18.04/x86_64
CPU : 8 Virtual CPUs
Memory : 16.0 GiB
# Go version and configurations
Go Version : go1.12.9 linux/amd64
GOMAXPROCS=8
# Machine information
OS : macOS Mojave 10.14.6/x86_64
CPU : 4 CPUs
Memory : 8.0 GiB
# Go version and configurations
Go Version : go version go1.12.9 darwin/amd64
GOMAXPROCS=4
Source code in gnet
is available under the MIT License.
Please read our Contributing Guidelines before opening a PR and thank you to all the developers who already made contributions to gnet
!
- A Million WebSockets and Go
- Going Infinite, handling 1M websockets connections in Go
- Go netpoll I/O 多路复用构建原生网络模型之源码深度解析
- gnet: 一个轻量级且高性能的 Golang 网络库
gnet
had been being developed with GoLand
IDE under the free JetBrains Open Source license(s) granted by JetBrains s.r.o., hence I would like to express my thanks here.
Support us with a monthly donation and help us continue our activities.
Become a bronze sponsor with a monthly donation of $10 and get your logo on our README on Github.