Table of contents
Directory
Start
TarsCPP
TarsJava
TarsGo
TarsPHP
Tars.js
Other Important Features
FAQ
Updated at 2020-03-20 16:00:22

Quick Start

Environment construction

Tars base component installation reference Deploy

Golang environment ready, tarsgo requires the golang version to be 1.9.x and above.。

Install tars: go get github.com/TarsCloud/TarsGo/tars

Compile the tars protocol to Golang tool:

cd $GOPATH/src/github.com/TarsCloud/TarsGo/tars/tools/tars2go && go build . 

cp tars2go $GOPATH/bin

Check whether the tars in the GOPATH path are successfully installed.

Service naming

Using the services of the Tars framework, the service name consists of three parts:

  • APP： The application name, which identifies a small collection of services. In the Tars system, the application name must be unique. For example: TestApp;

  • Server： Service name, the name of the process providing the service. The name of the server is named according to the service function. It is generally named: XXServer, for example, HelloServer;

  • Servant：A service provider that provides an interface or instance of a specific service. For example: HelloImp;

Instructions:

A Server can contain multiple Servant, and the system will use App + Server + Servant combination to define the routing name of the service in the system, called routing object. This name must be unique in the whole system, so that would uniquely identify itself when it externally served.

Therefore, when defining an App, you need pay attention to the uniqueness of the App.

For Example：TestApp.HelloServer.HelloObj.

Tars management system

When you login successfully, you will enter the Tars management system, as shown below:tars_manager_main

Under the menu tree of the Tars management system, the following functions are available:

  • Business management: Includes deployed services, service management, release management, service configuration, service monitoring, feature monitoring, etc.
  • Operation and maintenance: Includes service deployment, capacity expansion, template management, etc.

Service deployment

Service deployment can actually be done after service development, but it is recommended to do it first.

As shown below：

new_project

  • Application: Which application the service program belongs to, for example:TestApp。
  • Service Name: The identification name of the service program, for example: HelloGo。
  • Service type: In what language the service program is written, for example: tars_go.
  • Template: The name of the configuration file that is set when the service program is started. By default, tars.default can be used.
  • Node: the machine IP of the service deployment.
  • Set group: Set the Set group information of the service. The Set information includes 3 parts: Set name, Set area, and Set group name.
  • OBJ name: The name of Servant.
  • OBJ Binding IP: The machine IP of the service binding, generally the same as the node
  • Port: The port to which OBJ is to be bound.
  • Port type: Use TCP or UDP.
  • Protocol: The communication protocol used by the application layer. The Tars framework uses the tars protocol by default.
  • Number of threads: The number of business processing threads.
  • Maximum connections: The maximum number of connections supported.
  • Maximum queue length: The size of the request receiving queue.
  • Queue timeout: The timeout period for requesting the receive queue.

Click "Submit", after successful, the TestApp application under the menu number will appear the HelloServer name, and you will see the information of your new service program on the right side, as shown below:

service_inactive

Service development

Create service

Run the create_tars_server.sh script to automatically create the files necessary for the service, If there is a syntax error during execution, try to use dos2unix create_tars_server.sh to transcode.

sh $GOPATH/src/github.com/TarsCloud/TarsGo/tars/tools/create_tars_server.sh [App] [Server] [Servant]
E.g:  
sh $GOPATH/src/github.com/TarsCloud/TarsGo/tars/tools/create_tars_server.sh TestApp HelloGo SayHello

After the command is executed, the code will be generated into GOPATH, and the directory will be named in APP/Server. The generated code also prompts the specific path.

[root@1-1-1-1 ~]# sh $GOPATH/src/github.com/TarsCloud/TarsGo/tars/tools/create_tars_server.sh TestApp HelloGo SayHello
[create server: TestApp.HelloGo ...]
[mkdir: $GOPATH/src/TestApp/HelloGo/]
>>>Now doing:./start.sh >>>>
>>>Now doing:./Server.go >>>>
>>>Now doing:./Server.conf >>>>
>>>Now doing:./ServantImp.go >>>>
>>>Now doing:./makefile >>>>
>>>Now doing:./Servant.tars >>>>
>>>Now doing:client/client.go >>>>
>>>Now doing:vendor/vendor.json >>>>
# runtime/internal/sys
>>> Great！Done! You can jump in $GOPATH/src/TestApp/HelloGo
>>> Tip: After editing the JCE file, use the following to automatically generate the go file.
>>>       $GOPATH/bin/tars2go *.tars

Defining interface files

The interface file defines the request method and the parameter field and type. For the interface file definition rules, refer to tars_tup.md.

Just for test， we define an echoHello interface , the client request parameter is some short strings such as "tars", and the service will responds to client "hello tars".

# cat $GOPATH/src/TestApp/HelloGo/SayHello.tars 
module TestApp{
interface SayHello{
     int echoHello(string name, out string greeting); 
   };
};

Note: The out keyword in the parameter identifies the output parameter.

Server development

First convert the tars protocol file to the Golang language form

$GOPATH/bin/tars2go SayHello.tars

Now let's implement the logic of the server: the client sends a "name", and the server responds with the "hello name".

cat $GOPATH/src/TestApp/HelloGo/SayHelloImp.go

package main

type SayHelloImp struct {
}

func (imp *SayHelloImp) EchoHello(name string, greeting *string) (int32, error) {
     *greeting = "hello " + name
     return 0, nil
}

Note：The function name is capitalized here, and the Go language method need be exported.

Compile the main function, the initial code are generated by the tars framework.

cat $GOPATH/src/TestApp/HelloGo/HelloGo.go

package main

import (
	"github.com/TarsCloud/TarsGo/tars"

	"TestApp"
)

func main() { //Init servant
	imp := new(SayHelloImp)                                      //New Imp
	app := new(TestApp.SayHello)                                 //New init the A JCE
	cfg := tars.GetServerConfig()                               //Get Config File Object
	app.AddServant(imp, cfg.App+"."+cfg.Server+".SayHelloObj") //Register Servant
	tars.Run()
}

Compile and build the executable and package the release.

make && make tar

Will generate executable file HelloGo and release package HelloGo.tgz

Client development

package main

import (
        "fmt"
        "github.com/TarsCloud/TarsGo/tars"

        "TestApp"
)

var comm *tars.Communicator
func main() {
        comm = tars.NewCommunicator()
        obj := "TestApp.HelloGo.SayHelloObj@tcp -h 127.0.0.1 -p 3002 -t 60000"
        app := new(TestApp.SayHello)
        /*
         // if your service has been registered at tars registry
         comm = tars.NewCommunicator()
         obj := "TestApp.HelloGo.SayHelloObj"
         // tarsregistry service at 192.168.1.1:17890 
         comm.SetProperty("locator", "tars.tarsregistry.QueryObj@tcp -h 192.168.1.1 -p 17890")
        */
    
        comm.StringToProxy(obj, app)
        reqStr := "tars"
        var resp string
        ret, err := app.EchoHello(reqStr, &resp)
        if err != nil {
                fmt.Println(err)
                return
        }
        fmt.Println("ret: ", ret, "resp: ", resp)
}

  • TestApp dependencies code is generated by tars2go.。

  • obj specifies the server address port. If the server is not registered at tars registry, you need to know the address and port of the server and specify it in Obj. In the example, the protocol is TCP, the server address is local address, and the port is 3002. If you have multiple servant, you can write like this TestApp.HelloGo.SayHelloObj@tcp -h 127.0.0.1 -p 9985:tcp -h 192.168.1.1 -p 9983 so that requests can be spread across multiple servant.

    If the service has already been registered at tars registry, there is no need to write the server address and port, but you need to specify the address of the master when initializing the communicator.

  • com communicator, use communicating with the server.

Build and test

# go build client.go
# ./client
ret:  0 resp:  hello tars

HTTP service development

Tarsgo supports http service, follow the steps above to create a service, tarsgo processing http request is encapsulated in GO native net/http, so the usage is very simple.

package main

import (
	"net/http"
	"github.com/TarsCloud/TarsGo/tars"
)

func main() {
	mux := &tars.TarsHttpMux{}
	mux.HandleFunc("/", func(w http.ResponseWriter, r *http.Request) {
		w.Write([]byte("Hello tars"))
	})
        cfg := tars.GetServerConfig()
	tars.AddHttpServant(mux, cfg.App+"."+cfg.Server+".HttpSayHelloObj") //Register http server
	tars.Run()
}

In addition, you can directly call other tars services, the same as the "client development" mentioned.

Service release

In the menu tree of tars management system, find the service you deployed and click to enter the service page.

Select "Publish Management", select the node to be published, click "Publish Selected Node", click "Upload Release Package", select the released package, as shown below:

release

After uploading the distribution package, click on the "Select Release Version" drop-down box and the service program you uploaded will appear. Select the top one (latest uploaded).

Click "Publish", the service starts to be released, after the release is successful, the following interface appears, as shown below:

service_ok

Chapter