panjf2000
[Bug]: 运行正常情况下突然出现日志
Actions I've taken before I'm here
- [X] I've thoroughly read the documentations on this issue but still have no clue.
- [X] I've searched the current list of Github issues but didn't find any duplicate issues that have been solved.
- [X] I've searched the internet with this issue, but haven't found anything helpful.
What happened?
gnet 监听udp服务，打印日志[email protected]/reactor_default_linux.go:124 event-loop(6) is exiting due to error: server is going to be shutdown；随后进程终止
Major version of gnet
v2
Specific version of gnet
v2.1.2
Operating system
Linux
Relevant log output
[email protected]/reactor_default_linux.go:124 event-loop(6) is exiting due to error: server is going to be shutdown
Code snippets (optional)
No response
How to Reproduce
没特殊操作，正常运行的简单udp接收服务
Does this issue reproduce with the latest release?
It can reproduce with the latest release
有看到 error occurs in event-loop: 的错误日志吗？没有的话应该是你在 OnTraffic() 里 return Shutdown 导致的。
有看到
error occurs in event-loop:的错误日志吗？没有的话应该是你在 OnTraffic() 里 return Shutdown 导致的。
日志文件仅包含[email protected]/reactor_default_linux.go:124 event-loop(6) is exiting due to error: server is going to be shutdown
2023/01/12 16:48:45 Init config success. 这条日志。Ontraffic 无return shutdown
OnClose() 和 OnTick() 里有没有返回 Shutdown？或者其他可能的地方，你全局搜一下 Shutdown 看看？一般来说这个错误是人为触发的，你看看你自己的业务代码有没有哪里有错误日志？
日志都是正常的，就多了一行异常日志
OnClose() 和 OnTick() 里有没有返回 Shutdown？或者其他可能的地方，你全局搜一下
Shutdown看看？一般来说这个错误是人为触发的，你看看你自己的业务代码有没有哪里有错误日志？
有类似Kill命令会触发Shutdown吗
暂时想不到还有什么情况会出现这种情况。你能提供一下能复现这个问题的 demo 代码吗？
type netServer struct {
gnet.BuiltinEventEngine
eng gnet.Engine
network string
addr string
multicore bool
}
func (s *netServer) OnBoot(eng gnet.Engine) gnet.Action {
logger.Infof("running server on %s with multi-core=%t", fmt.Sprintf("%s://%s", s.network, s.addr), s.multicore)
s.eng = eng
return gnet.None
}
func (s *netServer) OnOpen(c gnet.Conn) ([]byte, gnet.Action) {
logger.Debugf("connected with fd: %d, remote_addr: %s\n", c.Fd(), c.RemoteAddr().String())
//c.SetContext(new(protocol.NetProtoCodec))
return nil, gnet.None
}
func (s *netServer) OnClose(c gnet.Conn, err error) (action gnet.Action) {
if err != nil {
logger.Errorf("error occurred on fd: %d, remote_addr: %s, %v\n", c.Fd(), c.RemoteAddr().String(), err)
}
return gnet.Close
}
func (s *netServer) OnTraffic(c gnet.Conn) (action gnet.Action) {
codec := new(NetProtoHeader)
data, err := codec.Decode(c)
if err != nil {
logger.Errorf("invalid packet: %v", err)
return gnet.Close
}
logger.Infof("gnet codec: %v", codec)
err = ants.Submit(func() {
switch codec.MsgType {
case MsgTypePTRQuery:
req := new(PTRReq)
err := req.Unpack(data)
if err != nil {
logger.Errorf("Query ptr invalid packet: %s", err.Error())
return
}
logger.Infof("PTR request: %+v", req)
payload, err := singlePTR(req)
if err != nil {
logger.Errorf("Query ptr record by ip %s failed: %s", req.Qip.IpAddr.String(), err.Error())
//return
header := NetProtoHeader{MsgType: MsgTypePTRRes}
resBody := DNSProtoSignalRes{
DstIp: req.DstIp,
ProtoBody: []byte{0x00, 0x00},
}
payload = header.Encode(resBody.Pack())
}
sendSignalReqChan <- signalReq{Dst: req.DstIp.IpAddr.String(), Type: "ptr response", Data: payload}
case MsgTypeTypeRes:
//ants.Submit(func(){
var ip net.IP
switch len(data) {
case 5 + 4, 17 + 4:
ip = data[1 : len(data)-4]
default:
logger.Errorf("invalid payload body length: %d", len(data))
return
}
logger.Infof("AIO addr: %s", ip.String())
key := fmt.Sprintf("%s,%d", ip.String(), MsgTypeTypeRes)
fn := func() (interface{}, error) {
cType := binary.BigEndian.Uint16(data[len(data)-4:])
ch, ok := ip2CType.Get(ip.String())
if !ok {
return nil, fmt.Errorf("invalid ip %s to conn type channel", ip.String())
}
ch <- cType
logger.Infof("Ip %s connection type %d", ip.String(), cType)
return nil, nil
}
_, err, _ := singleGroup.Do(key, fn)
if err != nil {
logger.Warnf("Handle ip %s type response failed %s", ip.String(), err.Error())
return
}
default:
logger.Errorf("invalid payload type: %d", codec.MsgType)
return
}
})
if err != nil {
logger.Errorf("Submits a task to ants pool failed %s", err.Error())
}
return
}
func InitPtrServer() {
port := viper.GetInt("ptr.port")
if port == 0 {
port = 30053
}
service := &netServer{
network: "udp",
addr: fmt.Sprintf(":%d", port),
multicore: true,
}
err := gnet.Run(service, service.network+"://"+service.addr, gnet.WithMulticore(service.multicore), gnet.WithLogger(logger))
if err != nil {
logger.Errorf("running server on %s with multi-core=%t failed", fmt.Sprintf("%s://%s", service.network, service.addr), service.multicore)
panic(err)
}
}
func InitPtrServer() {
port := viper.GetInt("ptr.port")
if port == 0 {
port = 30053
}
service := &netServer{
network: "udp",
addr: fmt.Sprintf(":%d", port),
multicore: true,
}
err := gnet.Run(service, service.network+"://"+service.addr, gnet.WithMulticore(service.multicore), gnet.WithLogger(logger))
if err != nil {
logger.Errorf("running server on %s with multi-core=%t failed", fmt.Sprintf("%s://%s", service.network, service.addr), service.multicore)
panic(err)
}
}
gnet服务启动是在一个协程内启动的，单独的gnet shutdown会导致整个进程断掉吗
什么意思？你是主动调用了 gnet.Stop 方法吗？
没主动调用 gnet.Stop方法，项目设计的gnet部分代码全部在上面了，上面的 InitPtrServer() 是 go InitPtrServer() 通过协程方式启动的，[email protected]/reactor_default_linux.go:124 event-loop(6) is exiting due to error: server is going to be shutdown 日志会导致整个进程断掉吗
gnet服务启动是在一个协程内启动的，单独的gnet shutdown会导致整个进程断掉吗
什么意思？你是主动调用了 gnet.Stop 方法吗？
没主动调用 gnet.Stop方法，项目设计的gnet部分代码全部在上面了，上面的 InitPtrServer() 是 go InitPtrServer() 通过协程方式启动的，[email protected]/reactor_default_linux.go:124 event-loop(6) is exiting due to error: server is going to be shutdown 日志会导致整个进程断掉吗
不会使得整个进程退出，只是使 gnet.Run() 结束而已，而且如果发生了 server is going to be shutdown 错误，应该也要打印所有 eventloop 的退出日志，而你这里就只打印了其中一个，这整个过程都太奇怪了，后来还有出现过相同的问题吗？
我准备优化一下错误打印这部分代码，把相应的堆栈信息也一起打印出来，这样后续定位问题更准确，至于你这个情况，麻烦你持续观察下，如果后面还有复现再一起来看看，谢谢。
🤖 Non-English text detected, translating...
It seems that this problem has not recurred for a long time, so we will temporarily close this issue. You can reopen it at any time if it happens again.
