When I send requests from the following code:
req, err := http.NewRequest("GET", "my_target:", nil)
if err != nil {
panic(err)
}
req.Close = true
client := http.DefaultClient
resp, err := client.Do(req)
if err != nil {
panic(err)
}
defer resp.Body.Close()
After a few hours of sending average 10 requests per minutes, I get this error:
socket: too many open files
How do I find the cause of this problem?
Am I not closing the http.Request?
I thought req.Close = true does that job.
Thanks!
Why are you deferring the close? Are you actually reading from this body?
defer resp.Body.Close()
Do you actually return from the current function before performing another Get? If not, then the defer will never execute, and you'll never release this connection for reuse.
req.Close = true is an unusual choice here, as well. This also prevents connection reuse, which is something you'd probably want rather than forbid. This doesn't automatically close the request on your side. It forces the server to immediately close the connection, which you would otherwise reuse. You'll hold your side open until you close it.
Typically for a simple GET request like you have here, I would just do this:
resp, err := http.Get("...")
if err != nil {
panic(err) // panic seems harsh, usually you'd just return the error. But either way...
}
resp.Body.Close()
There's no need for a special client here. Just use the default one. It'll take care of the rest as long as you make sure to close the response body. And there's no need to complicate things with a defer if you're going to immediately close the body. The reason for defer is to make sure you close the body if you have a bunch of processing later that might return or panic on error.
Related
As per the documentation in Readme:
Make sure to defer a call to Disconnect after instantiating your client:
defer func() {
if err = client.Disconnect(ctx); err != nil {
panic(err)
}
}()
Does the above documentation meant to disconnect, during shutdown of the program(using mongodb driver)?
They just reminded you that you should always close the connection to the database at some point. When exactly is up to you. Usually, you initialize the database connection at the top-level of your application, so the defer call should be at the same level. For example,
func main() {
client, err := mongo.Connect(ctx, options.Client().ApplyURI("mongodb://localhost:27017"))
defer func() {
if err = client.Disconnect(ctx); err != nil {
panic(err)
}
}()
// Pass the connection to other components of your appliaction
doSomeWork(client)
}
Note: If you are using Goroutines don't forget to synchronize them in main otherwise the connection will be close too early.
I think the documentation meant to use defer with client.Disconnect in the main function of your program. Thanks to that your program will close all the MongoDB client connections before exiting.
If you would use it e.g. in a helper function that prepares that client, it would close all the connections right after the client creation which may not be something you want.
How do I create a proper mongo based application in Go using the official driver(go.mongodb.org/mongo-driver/mongo)? I have a MongoConnect() and MongoDisconnect(client) function to create a connection and delete it. But, it's not too efficient and starts leaking FD as the app has got around 40 functions and finding all the missed MongoDisconnect() becomes hectic.
The current MongoConnect and MongoDisconnect are as follows.
func MongoConnect() (*mongo.Client, error) {
clientOptions := options.Client().ApplyURI("mongodb://localhost:27017")
client, err := mongo.Connect(context.TODO(), clientOptions)
if err != nil {
Logger(err)
return nil, err
}
err = client.Ping(context.TODO(), nil)
if err != nil {
Logger(err)
return nil, err
}
return client, err
}
func MongoDisconnect(client *mongo.Client) {
_ = client.Disconnect(context.TODO())
}
I am looking for a method that would still use MongoConnect() to create connections and would automatically kill the client without the usage of MongoDisconnect().
PS. Other methods that are better than the above requirement are also welcome
I'm not sure that there is an 'efficient' way to fix the underlying issue, you probably will need to look at all the places where you've called MongoConnect() and ensure you have a corresponding MongoDisconnect().
In saying that, what you might want to look at is implemententing the right pattern for connecting to databases.
Generally speaking, if your database driver takes care of managing connections for you then you should create the connection once and just pass it around as needed.
You could also defer the closing of that connection to a go routine which would close it once it was no longer needed (when you're application is shutting down).
Here is a code snippet of how this is implemented:
// =========================================================================
// Start Database
log.Println("main: Initializing database support")
db, err := database.Open(database.Config{
User: cfg.DB.User,
Password: cfg.DB.Password,
Host: cfg.DB.Host,
Name: cfg.DB.Name,
DisableTLS: cfg.DB.DisableTLS,
})
if err != nil {
return errors.Wrap(err, "connecting to db")
}
defer func() {
log.Printf("main: Database Stopping : %s", cfg.DB.Host)
db.Close()
}()
I didn't write this code and its part of a larger project scaffold which has some other nice patterns for web applications.
Ultimate service
I have a websocket client. In reality, it is far more complex than the basic code shown below.
I now need to scale this client code to open connections to multiple servers. Ultimately, the tasks that need to be performed when a message is received from the servers is identical.
What would be the best approach to handle this?
As I said above the actual code performed when receiving the message is far more complex than shown in the example.
package main
import (
"flag"
"log"
"net/url"
"os"
"os/signal"
"time"
"github.com/gorilla/websocket"
)
var addr = flag.String("addr", "localhost:1234", "http service address")
func main() {
flag.Parse()
log.SetFlags(0)
interrupt := make(chan os.Signal, 1)
signal.Notify(interrupt, os.Interrupt)
// u := url.URL{Scheme: "ws", Host: *addr, Path: "/echo"}
u := url.URL{Scheme: "ws", Host: *addr, Path: "/"}
log.Printf("connecting to %s", u.String())
c, _, err := websocket.DefaultDialer.Dial(u.String(), nil)
if err != nil {
log.Fatal("dial:", err)
}
defer c.Close()
done := make(chan struct{})
go func() {
defer close(done)
for {
_, message, err := c.ReadMessage()
if err != nil {
log.Println("read:", err)
return
}
log.Printf("recv: %s", message)
}
}()
ticker := time.NewTicker(time.Second)
defer ticker.Stop()
for {
select {
case <-done:
return
case t := <-ticker.C:
err := c.WriteMessage(websocket.TextMessage, []byte(t.String()))
if err != nil {
log.Println("write:", err)
return
}
case <-interrupt:
log.Println("interrupt")
// Cleanly close the connection by sending a close message and then
// waiting (with timeout) for the server to close the connection.
err := c.WriteMessage(websocket.CloseMessage, websocket.FormatCloseMessage(websocket.CloseNormalClosure, ""))
if err != nil {
log.Println("write close:", err)
return
}
select {
case <-done:
case <-time.After(time.Second):
}
return
}
}
}
Modify the interrupt handling to close a channel on interrupt. This allows multiple goroutines to wait on the event by waiting for the channel to close.
shutdown := make(chan struct{})
interrupt := make(chan os.Signal, 1)
signal.Notify(interrupt, os.Interrupt)
go func() {
<-interrupt
log.Println("interrupt")
close(shutdown)
}()
Move the per-connection code to a function. This code is a copy and paste from the question with two changes: the interrupt channel is replaced with the shutdown channel; the function notifies a sync.WaitGroup when the function is done.
func connect(u string, shutdown chan struct{}, wg *sync.WaitGroup) {
defer wg.Done()
log.Printf("connecting to %s", u)
c, _, err := websocket.DefaultDialer.Dial(u, nil)
if err != nil {
log.Fatal("dial:", err)
}
defer c.Close()
done := make(chan struct{})
go func() {
defer close(done)
for {
_, message, err := c.ReadMessage()
if err != nil {
log.Println("read:", err)
return
}
log.Printf("recv: %s", message)
}
}()
ticker := time.NewTicker(time.Second)
defer ticker.Stop()
for {
select {
case <-done:
return
case t := <-ticker.C:
err := c.WriteMessage(websocket.TextMessage, []byte(t.String()))
if err != nil {
log.Println("write:", err)
return
}
case <-shutdown:
// Cleanly close the connection by sending a close message and then
// waiting (with timeout) for the server to close the connection.
err := c.WriteMessage(websocket.CloseMessage, websocket.FormatCloseMessage(websocket.CloseNormalClosure, ""))
if err != nil {
log.Println("write close:", err)
return
}
select {
case <-done:
case <-time.After(time.Second):
}
return
}
}
}
Declare a sync.WaitGroup in main(). For each websocket endpoint that you want to connect to, increment the WaitGroup and start a goroutine to connect that endpoint. After starting the goroutines, wait on the WaitGroup for the goroutines to complete.
var wg sync.WaitGroup
for _, u := range endpoints { // endpoints is []string
// where elements are URLs
// of endpoints to connect to.
wg.Add(1)
go connect(u, shutdown, &wg)
}
wg.Wait()
The code above with an edit to make it run against Gorilla's echo example server is posted on the playground.
is the communication with every different server completely independendant of the other servers? if yes i would go around in a fashion like:
in main create a context with a cancellation function
create a waitgroup in main to track fired up goroutines
for every server, add to the waitgroup, fire up a new goroutine from the main function passing the context and the waitgroup references
main goes in a for/select loop listening to for signals and if one arrives calls the cancelfunc and waits on the waitgroup.
main can also listen on a result chan from the goroutines and maybe print the results itself it the goroutines shouldn't do it directly.
every goroutine has as we said has references for the wg, the context and possibly a chan to return results. now the approach splits on if the goroutine must do one and one thing only, or if it needs to do a sequence of things. for the first approach
if only one thing is to be done we follow an approach like the one descripbed here (observe that to be asyncronous he would in turn fire up a new goroutine to perform the DoSomething() step that would return the result on the channel)
That allows it to be able to accept the cancellation signal at any time. it is up to you to determine how non-blocking you want to be and how prompt you want to be to respond to cancellation signals.Also the benefit of having the a context associated being passed to the goroutines is that you can call the Context enabled versions of most library functions. For example if you want your dials to have a timeout of let's say 1 minute, you would create a new context with timeout from the one passed and then DialContext with that. This allows the dial to stop both from a timeout or the parent (the one you created in main) context's cancelfunc being called.
if more things need to be done ,i usually prefer to do one thing with the goroutine, have it invoke a new one with the next step to be performed (passing all the references down the pipeline) and exit.
this approach scales well with cancellations and being able to stop the pipeline at any step as well as support contexts with dealines easily for steps that can take too long.
Talk is cheap, so here we go the simple code:
package main
import (
"fmt"
"time"
"net"
)
func main() {
addr := "127.0.0.1:8999"
// Server
go func() {
tcpaddr, err := net.ResolveTCPAddr("tcp4", addr)
if err != nil {
panic(err)
}
listen, err := net.ListenTCP("tcp", tcpaddr)
if err != nil {
panic(err)
}
for {
if conn, err := listen.Accept(); err != nil {
panic(err)
} else if conn != nil {
go func(conn net.Conn) {
buffer := make([]byte, 1024)
n, err := conn.Read(buffer)
if err != nil {
fmt.Println(err)
} else {
fmt.Println(">", string(buffer[0 : n]))
}
conn.Close()
}(conn)
}
}
}()
time.Sleep(time.Second)
// Client
if conn, err := net.Dial("tcp", addr); err == nil {
for i := 0; i < 2; i++ {
_, err := conn.Write([]byte("hello"))
if err != nil {
fmt.Println(err)
conn.Close()
break
} else {
fmt.Println("ok")
}
// sleep 10 seconds and re-send
time.Sleep(10*time.Second)
}
} else {
panic(err)
}
}
Ouput:
> hello
ok
ok
The Client writes to the Server twice. After the first read, the Server closes the connection immediately, but the Client sleeps 10 seconds and then re-writes to the Server with the same already closed connection object(conn).
Why can the second write succeed (returned error is nil)?
Can anyone help?
PS:
In order to check if the buffering feature of the system affects the result of the second write, I edited the Client like this, but it still succeeds:
// Client
if conn, err := net.Dial("tcp", addr); err == nil {
_, err := conn.Write([]byte("hello"))
if err != nil {
fmt.Println(err)
conn.Close()
return
} else {
fmt.Println("ok")
}
// sleep 10 seconds and re-send
time.Sleep(10*time.Second)
b := make([]byte, 400000)
for i := range b {
b[i] = 'x'
}
n, err := conn.Write(b)
if err != nil {
fmt.Println(err)
conn.Close()
return
} else {
fmt.Println("ok", n)
}
// sleep 10 seconds and re-send
time.Sleep(10*time.Second)
} else {
panic(err)
}
And here is the screenshot:
attachment
There are several problems with your approach.
Sort-of a preface
The first one is that you do not wait for the server goroutine
to complete.
In Go, once main() exits for whatever reason,
all the other goroutines still running, if any, are simply
teared down forcibly.
You're trying to "synchronize" things using timers,
but this only works in toy situations, and even then it
does so only from time to time.
Hence let's fix your code first:
package main
import (
"fmt"
"log"
"net"
"time"
)
func main() {
addr := "127.0.0.1:8999"
tcpaddr, err := net.ResolveTCPAddr("tcp4", addr)
if err != nil {
log.Fatal(err)
}
listener, err := net.ListenTCP("tcp", tcpaddr)
if err != nil {
log.Fatal(err)
}
// Server
done := make(chan error)
go func(listener net.Listener, done chan<- error) {
for {
conn, err := listener.Accept()
if err != nil {
done <- err
return
}
go func(conn net.Conn) {
var buffer [1024]byte
n, err := conn.Read(buffer[:])
if err != nil {
log.Println(err)
} else {
log.Println(">", string(buffer[0:n]))
}
if err := conn.Close(); err != nil {
log.Println("error closing server conn:", err)
}
}(conn)
}
}(listener, done)
// Client
conn, err := net.Dial("tcp", addr)
if err != nil {
log.Fatal(err)
}
for i := 0; i < 2; i++ {
_, err := conn.Write([]byte("hello"))
if err != nil {
log.Println(err)
err = conn.Close()
if err != nil {
log.Println("error closing client conn:", err)
}
break
}
fmt.Println("ok")
time.Sleep(2 * time.Second)
}
// Shut the server down and wait for it to report back
err = listener.Close()
if err != nil {
log.Fatal("error closing listener:", err)
}
err = <-done
if err != nil {
log.Println("server returned:", err)
}
}
I've spilled a couple of minor fixes
like using log.Fatal (which is
log.Print + os.Exit(1)) instead of panicking,
removed useless else clauses to adhere to the coding standard of keeping the main
flow where it belongs, and lowered the client's timeout.
I have also added checking for possible errors Close on sockets may return.
The interesting part is that we now properly shut the server down by closing the listener and then waiting for the server goroutine to report back (unfortunately Go does not return an error of a custom type from net.Listener.Accept in this case so we can't really check that Accept exited because we've closed the listener).
Anyway, our goroutines are now properly synchronized, and there is
no undefined behaviour, so we can reason about how the code works.
Remaining problems
Some problems still remain.
The more glaring is you making wrong assumption that TCP preserves
message boundaries—that is, if you write "hello" to the client
end of the socket, the server reads back "hello".
This is not true: TCP considers both ends of the connection
as producing and consuming opaque streams of bytes.
This means, when the client writes "hello", the client's
TCP stack is free to deliver "he" and postpone sending "llo",
and the server's stack is free to yield "hell" to the read
call on the socket and only return "o" (and possibly some other
data) in a later read.
So, to make the code "real" you'd need to somehow introduce these
message boundaries into the protocol above TCP.
In this particular case the simplest approach would be either
using "messages" consisting of a fixed-length and agreed-upon
endianness prefix indicating the length of the following
data and then the string data itself.
The server would then use a sequence like
var msg [4100]byte
_, err := io.ReadFull(sock, msg[:4])
if err != nil { ... }
mlen := int(binary.BigEndian.Uint32(msg[:4]))
if mlen < 0 {
// handle error
}
if mlen == 0 {
// empty message; goto 1
}
_, err = io.ReadFull(sock, msg[5:5+mlen])
if err != nil { ... }
s := string(msg[5:5+mlen])
Another approach is to agree on that the messages do not contain
newlines and terminate each message with a newline
(ASCII LF, \n, 0x0a).
The server side would then use something like
a usual bufio.Scanner loop to get
full lines from the socket.
The remaining problem with your approach is to not dealing with
what Read on a socket returns: note that io.Reader.Read
(that's what sockets implement, among other things) is allowed
to return an error while having had read some data from the
underlying stream. In your toy example this might rightfully
be unimportant, but suppose that you're writing a wget-like
tool which is able to resume downloading of a file: even if
reading from the server returned some data and an error, you
have to deal with that returned chunk first and only then
handle the error.
Back to the problem at hand
The problem presented in the question, I beleive, happens simply because in your setup you hit some TCP buffering problem due to the tiny length of your messages.
On my box which runs Linux 4.9/amd64 two things reliably "fix"
the problem:
Sending messages of 4000 bytes in length: the second call
to Write "sees" the problem immediately.
Doing more Write calls.
For the former, try something like
msg := make([]byte, 4000)
for i := range msg {
msg[i] = 'x'
}
for {
_, err := conn.Write(msg)
...
and for the latter—something like
for {
_, err := conn.Write([]byte("hello"))
...
fmt.Println("ok")
time.Sleep(time.Second / 2)
}
(it's sensible to lower the pause between sending stuff in
both cases).
It's interesting to note that the former example hits the
write: connection reset by peer (ECONNRESET in POSIX)
error while the second one hits write: broken pipe
(EPIPE in POSIX).
This is because when we're sending in chunks worth 4k bytes,
some of the packets generated for the stream manage to become
"in flight" before the server's side of the connection manages
to propagate the information on its closure to the client,
and those packets hit an already closed socket and get rejected
with the RST TCP flag set.
In the second example an attempt to send another chunk of data
sees that the client side already knows that the connection
has been teared down and fails the sending without "touching
the wire".
TL;DR, the bottom line
Welcome to the wonderful world of networking. ;-)
I'd recommend buying a copy of "TCP/IP Illustrated",
read it and experiment.
TCP (and IP and other protocols above IP)
sometimes works not like people expect them to by applying
their "common sense".
So I'm having some trouble figuring out best practices for using concurrency with a MongoDB in go. My first implementation of getting a session looked like this:
var globalSession *mgo.Session
func getSession() (*mgo.Session, error) {
//Establish our database connection
if globalSession == nil {
var err error
globalSession, err = mgo.Dial(":27017")
if err != nil {
return nil, err
}
//Optional. Switch the session to a monotonic behavior.
globalSession.SetMode(mgo.Monotonic, true)
}
return globalSession.Copy(), nil
}
This works great the trouble I'm running into is that mongo has a limit of 204 connections then it starts refusing connections connection refused because too many open connections: 204;however, the issue is since I'm calling session.Copy() it only returns a session and not an error. So event though the connection refused my program never thrown an error.
Now what I though about doing is just having one session and using that instead of copy so I can have access to a connection error like so:
var session *mgo.Session = nil
func NewSession() (*mgo.Session, error) {
if session == nil {
session, err = mgo.Dial(url)
if err != nil {
return nil, err
}
}
return session, nil
}
Now the problem I have with this is that I don't know what would happen if I try to make concurrent usage of that same session.
The key is to duplicate the session and then close it when you've finished with it.
func GetMyData() []myMongoDoc {
sessionCopy, _ := getSession() // from the question above
defer sessionCopy.Close() // this is the important bit
results := make([]myMongoDoc, 0)
sessionCopy.DB("myDB").C("myCollection").Find(nil).All(&results)
return results
}
Having said that it looks like mgo doesn't actually expose control over the underlying connections (see the comment from Gustavo Niemeyer who maintains the library). A session pretty much equates to a connection, but even if you call Close() on a session mgo keeps the connection alive. From reading around it seems that Clone() might be the way to go, as it reuses the underlying socket, this will avoid the 3 way handshake of creating a new socket (see here for more discussion on the difference).
Also see this SO answer describing a standard pattern to handle sessions.