Skip to content

aliyun/aliyun-log-go-sdk

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

User Guide (中文)

README in English

基本介绍

本项目是阿里云日志服务 (Log Service,简称SLS)API的Golang编程接口,提供了对于Log Service Rest API的封装和实现,帮助Golang开发人员更快编程使用阿里云日志服务。

本项目主要由3个部分组成:

  1. 日志服务基础API封装和实现。
  2. Golang Producer Library,用于向日志服务批量发送数据,详情参考Aliyun LOG Golang Producer 快速入门
  3. Golang Consumer Library,用于消费日志服务中的数据,详情参考Consumer Library

详细API接口以及含义请参考:https://help.aliyun.com/document_detail/29007.html

安装

go get -u github.com/aliyun/aliyun-log-go-sdk

快速入门

前言: 所有的使用样例都位于example目录下,使用该目录下的所有样例前,请先在该目录下的config.go文件中的init函数中配置您的 project, logstore等所需要的配置参数,example目录下的所有样例都会使用config.go文件中配置的参数。

  1. 创建Client

    参考 config.go 文件

    AccessKeyID = "your ak id"
    AccessKeySecret = "your ak secret"
    credentialsProvider := NewStaticCredentialsProvider(AccessKeyID, AccessKeySecret, "")
    Endpoint = "your endpoint" // just like cn-hangzhou.log.aliyuncs.com
    Client = sls.CreateNormalInterfaceV2(Endpoint, credentialsProvider)

    为了防止出现配置错误,您可以在创建 Client 之后,测试 Client 是否能成功调用 SLS API

    _, err := Client.ListProject()
    if err != nil {
       panic(err)
    }
  2. 创建project

    参考 log_project.go文件

    project, err := util.Client.CreateProject(ProjectName,"Project used for testing")
    if err != nil {
       fmt.Println(err)
    }
    fmt.Println(project)
  3. 创建logstore

    参考 log_logstore.go

    err := util.Client.CreateLogStore(ProjectName, LogStoreName,2,2,true,64)
    if err != nil {
       fmt.Println(err)
    }
  4. 创建索引

    参考index_sample

    indexKeys := map[string]sls.IndexKey{
       "col_0": {
          Token:         []string{" "},
          CaseSensitive: false,
          Type:          "long",
       },
       "col_1": {
          Token:         []string{",", ":", " "},
          CaseSensitive: false,
          Type:          "text",
       },
    }
    index := sls.Index{
       Keys: indexKeys,
       Line: &sls.IndexLine{
          Token:         []string{",", ":", " "},
          CaseSensitive: false,
          IncludeKeys:   []string{},
          ExcludeKeys:   []string{},
       },
    }
    err = util.Client.CreateIndex(util.ProjectName, logstore_name, index)
    if err != nil {
       fmt.Printf("CreateIndex fail, err: %s", err)
       return
    }
    fmt.Println("CreateIndex success")
  5. 写数据

    这里展示了用sdk中原生的API接口去发送数据简单示例,但是我们不推荐用API直接向logstore写入数据,推荐使用SDK 中提供的producer 包向logstore 写入数据,自动压缩数据并且提供安全退出机制,不会使数据丢失, 对于MetricStore类型, 使用PutLogsWithMetricStoreURLAPI 发送数据可以提升大基数时间线下查询性能 。

    logs := []*sls.Log{}
    for logIdx := 0; logIdx < 100; logIdx++ {
       content := []*sls.LogContent{}
       for colIdx := 0; colIdx < 10; colIdx++ {
          content = append(content, &sls.LogContent{
             Key:   proto.String(fmt.Sprintf("col_%d", colIdx)),
             Value: proto.String(fmt.Sprintf("loggroup idx: %d, log idx: %d, col idx: %d, value: %d", loggroupIdx, logIdx, colIdx, rand.Intn(10000000))),
          })
       }
       log := &sls.Log{
          Time:     proto.Uint32(uint32(time.Now().Unix())),
          Contents: content,
       }
       logs = append(logs, log)
    }
    loggroup := &sls.LogGroup{
       Topic:  proto.String(""),
       Source: proto.String("10.230.201.117"),
       Logs:   logs,
    }
    // PostLogStoreLogs API Ref: https://intl.aliyun.com/help/doc-detail/29026.htm
    for retryTimes := 0; retryTimes < 10; retryTimes++ {
       err := util.Client.PutLogs(util.ProjectName, util.LogStoreName, loggroup)
       if err == nil {
          fmt.Printf("PutLogs success, retry: %d\n", retryTimes)
          break
       } else {
          fmt.Printf("PutLogs fail, retry: %d, err: %s\n", retryTimes, err)
          //handle exception here, you can add retryable erorrCode, set appropriate put_retry
          if strings.Contains(err.Error(), sls.WRITE_QUOTA_EXCEED) || strings.Contains(err.Error(), sls.PROJECT_QUOTA_EXCEED) || strings.Contains(err.Error(), sls.SHARD_WRITE_QUOTA_EXCEED) {
             //mayby you should split shard
             time.Sleep(1000 * time.Millisecond)
          } else if strings.Contains(err.Error(), sls.INTERNAL_SERVER_ERROR) || strings.Contains(err.Error(), sls.SERVER_BUSY) {
             time.Sleep(200 * time.Millisecond)
          }
       }
    }

6.读数据

这里展示了使用SDK中原生API接口调用去拉取数据的方式,我们不推荐使用这种方式去读取消费logstore中的数据,推荐使用SDK中 consumer 消费组去拉取数据,消费组提供自动负载均衡以及失败重试等机制,并且会自动保存拉取断点,再次拉取不会拉取重复数据。

shards, err := client.ListShards(project, logstore)

totalLogCount := 0
	for _, shard := range shards {
		fmt.Printf("[shard] %d begin\n", shard.ShardID)
		beginCursor, err := client.GetCursor(project, logstore, shard.ShardID, "begin")
		if err != nil {
			panic(err)
		}
		endCursor, err := client.GetCursor(project, logstore, shard.ShardID, "end")
		if err != nil {
			panic(err)
		}

		nextCursor := beginCursor
		for nextCursor != endCursor {
			gl, nc, err := client.PullLogs(project, logstore, shard.ShardID, nextCursor, endCursor, 10)
			if err != nil {
				fmt.Printf("pull log error : %s\n", err)
				time.Sleep(time.Second)
				continue
			}
			nextCursor = nc
			fmt.Printf("now count %d \n", totalLogCount)
			if gl != nil {
				for _, lg := range gl.LogGroups {
					for _, tag := range lg.LogTags {
						fmt.Printf("[tag] %s : %s\n", tag.GetKey(), tag.GetValue())
					}
					for _, log := range lg.Logs {
						totalLogCount++
						// print log
						for _, content := range log.Contents {
							fmt.Printf("[log] %s : %s\n", content.GetKey(), content.GetValue())
						}
					}
				}
			}
		}
  1. 创建机器组

    参考 machine_group_sample.go

机器组分为IP型机器组和标识型机器组,二选一

创建标识型机器组

attribute := sls.MachinGroupAttribute{
}
machineList := []string{"mac-user-defined-id-value_XX"}
var machineGroup = &sls.MachineGroup{
   Name:          groupName,
   MachineIDType: "userdefined",
   MachineIDList: machineList,
   Attribute:     attribute,
}
err = util.Client.CreateMachineGroup(ProjectName, machineGroup)
if err != nil {
   fmt.Println(err)
}

创建IP型机器组

attribute := sls.MachinGroupAttribute{
}
machineList := []string{"192.168.XX.XX"}
var machineGroup = &sls.MachineGroup{
   Name:          groupName,
   MachineIDType: "ip",
   MachineIDList: machineList,
   Attribute:     attribute,
}
err = util.Client.CreateMachineGroup(ProjectName, machineGroup)
if err != nil {
   fmt.Println(err)
}
  1. 创建logtail 采集配置

    logtail 采集配置,目前通过sdk 支持创建下列几种模式的采集配置,分别为 完整正则分隔符模式json模式插件模式,这里展示的完整正则模式的创建。

    regexConfig := new(sls.RegexConfigInputDetail)
    regexConfig.DiscardUnmatch = false
    regexConfig.Key = []string{"logger", "time", "cluster", "hostname", "sr", "app", "workdir", "exe", "corepath", "signature", "backtrace"}
    regexConfig.Regex = "\\S*\\s+(\\S*)\\s+(\\S*\\s+\\S*)\\s+\\S*\\s+(\\S*)\\s+(\\S*)\\s+(\\S*)\\s+(\\S*)\\s+(\\S*)\\s+(\\S*)\\s+(\\S*)\\s+\\S*\\s+(\\S*)\\s*([^$]+)"
    regexConfig.TimeFormat = "%Y/%m/%d %H:%M:%S"
    regexConfig.LogBeginRegex = `INFO core_dump_info_data .*`
    regexConfig.LogPath = "/cloud/log/tianji/TianjiClient#/core_dump_manager"
    regexConfig.FilePattern = "core_dump_info_data.log*"
    regexConfig.MaxDepth = 0
    sls.InitRegexConfigInputDetail(regexConfig)
    outputDetail := sls.OutputDetail{
       ProjectName:  projectName,
       LogStoreName: logstore,
    }
    logConfig := &sls.LogConfig{
       Name:         configName,
       InputType:    "file",
       OutputType:   "LogService", // Now only supports LogService
       InputDetail:  regexConfig,
       OutputDetail: outputDetail,
    }
    err = util.Client.CreateConfig(projectName, logConfig)
    if err != nil {
       fmt.Println(err)
    }
  2. logtail配置到机器组

    err = util.Client.ApplyConfigToMachineGroup(ProjectName, confName, mgname)
    if err != nil {
       fmt.Println(err)
    }

开发者

更新protobuf工具

protoc -I=. -I=$GOPATH/src -I=$GOPATH/src/github.com/gogo/protobuf/protobuf --gofast_out=. log.proto