标签: aws-step-functions

文档定义了endpoint如此:

endpoint(String) - 发送请求的端点URI.默认端点是根据配置构建的region.端点应该是一个字符串'https://{service}.{region}.amazonaws.com'.

我相信服务名称是StepFunctions,我正在使用该区域us-east-1,因此网址应该是:

https://stepfunctions.us-east-1.amazonaws.com

但是,这似乎不对,这项服务的实际端点网址是什么？

编辑:(回答)

https://states.us-east-1.amazonaws.com

我有一个步骤函数，它应该调用 API 网关资源而不是 lambda。这样做的语法是什么？

{"Comment": "A Hello World example of the Amazon States Language using a Pass state",
  "StartAt": "QueueProducts",
  "States": {
    "GetProductsFromDb": {
      "Type": "Task",
      "Resource":"some-lambda",
      "Next": "InvokeAPIGatewayWorkers"
    }
  },
 "InvokeAPIGatewayWorkers":{
    "Type": "Parallel",
    "Branches": [
     ....]
}
}

我的问题是，是否可以在资源中调用 API 网关而不是“some-lamda”

I have a parallel task in the step function that contains two branches. The input was:

{
  "database": "test",
  "userName": "tester",
  "userID": "test123",
  "adGroup": "testADGroup",
  "dbGroup": "ReadGroup"
}

Each branch return a json result like the following

Branch 1 (I used "OutputPath": "$"):

{
  "requestType": "GrantAccess",
  "DBUser": "exists",
  "ADUser": "exists"
}

Branch 2 (I used "ResultPath": "$.approvalStatus"):

{
      "database": "test",
      "userName": "tester",
      "userID": "test123",
      "adGroup": "testADGroup",
      "dbGroup": "ReadGroup"
      "approvalStatus": "Approved"
}

When both the branches complete, the output of the …

我正在考虑使用步骤函数每 30 秒轮询一次来自外部数据库的更新，并运行映射 lambda 将数据存储在 s3 中。

实现步骤函数非常简单。我注意到限制的有效期为一年。http://docs.aws.amazon.com/step-functions/latest/dg/limits.html

我的问题是这是否有效，或者为此目的使用步骤函数是否存在缺陷。对于我的场景来说，每 1000 步 0.025 的价格是可以接受的。

我有以下步骤功能。执行在选择状态下失败。

"States": {
    "Process": {
        "Type": "Task",
        "Resource": "arn:aws:lambda:us-east-1:123:function:dummy1",
        "OutputPath": "$",
        "Next": "ChoiceStatePre"
    },

    "ChoiceStatePre": {
        "Type": "Choice",
        "Choices": [{
            "Variable": "$.status_pre",
            "Next": "Series0",
            "NumericEquals": 1
        }, {
            "Variable": "$.status_pre",
            "Next": "MatchStatePre",
            "NumericEquals": 8
        }]
    },
    "MatchStatePre": {
        "Type": "Task",
        "Next": "PreProcess",
        "Resource": "arn:aws:states:us-east-1:123:activity:dummy2"
    },
    "Series0": {
        "Resource": "arn:aws:lambda:us-east-1:123:function:dummy3",
        "Type": "Task",
        "InputPath": "$.seq0.step0",
        "ResultPath": "$.seq0.step0",
        "OutputPath": "$",
        "Next": "ChoiceStateTrigger0"
    }
}

错误 -

{
"error": "States.Runtime",
"cause": "An error occurred while executing the state 'ChoiceStatePre' 
(entered at the event id …

我想使用 Terraforms api_gateway_integration和sfn_state_machine使用 API 网关创建一个步进函数 API。

我现在必须在 api_gateway_integration 处填写uri 参数。

我的 step 函数已创建，我可以引用 step 函数的 id（类似于arn:aws:states:*region*:*account*:stateMachine:*step-function-name*:stateMachine:*step-function-entry-point*）。

如果 AWS 阶跃函数是目标，谁能告诉我该方案或 uri 参数外观的示例？

resource "aws_api_gateway_integration" "endpoint_integration" {
  ...
  integration_http_method = "POST"
  type = "AWS"
  uri = <<<<< What to place here???
}

我正在研究AWS步骤功能，但是无法找到用例需要的问题的答案。假设我们总体上可以完成10个任务。有时，我们必须按顺序执行任务1、3，有时执行4、8，而有时执行1、9、5。

要执行的任务及其执行顺序由传入json数据到系统中确定。

我想知道是否可以根据出现的需求动态创建步进功能。

我正在尝试创建一个可以调用另一个状态机的状态机。我尝试使用以下方法来获取 ARN。但是，这会返回错误 Arn is not a valid property，正在创建哪个堆栈。

  ParentStateMachine:
    Type: "AWS::StepFunctions::StateMachine"
    Properties:
      StateMachineName: !Sub "ParentStateMachine"
      DefinitionString:
        Fn::Sub:
         - |-
            {
              "Comment": "...",
              "StartAt": "State1",
              "States": {
                "State1": {
                  "Type": "Task",
                  "Resource": "arn:aws:states:::states:startExecution.sync",
                  "Parameters": {
                    "StateMachineArn": "${ChildStateMachineArn}",
                    "Input": {
                      "StatePayload": {
                        "datasetDate.$": "$.datasetDate"
                      },
                      "AWS_STEP_FUNCTIONS_STARTED_BY_EXECUTION_ID.$": "$$.Execution.Id"
                    }
                  },
                  "End": true
                }
              }
            }
         -  {
               ChildStateMachineArn:
                 Fn::GetAtt:
                   - ChildStateMachine
                   - Arn
            }
      RoleArn:
        Fn::GetAtt:
          - StatesExecutionRole
          - Arn

我还尝试使用此字符串生成 ARN。

arn:aws:states:${AWS::Region}:${AWS::AccountId}:stateMachine:ChildStateMachine

但是，这给出了错误

Failed to call Step Functions for request: 'com.amazonaws.services.stepfunctions.model.CreateStateMachineRequest'. (Service: …

我创建了 3 个 Glue 作业，其中作业参数键/值之一如下所示： runid id

如果我像这样使用 AWS CLI 执行 Glue 作业，则它工作正常： awsglue start-job-run --jobname $job --arguments='--runid="Runid_10"'

这 3 个 Glue 作业位于一步函数和状态机内，定义为：

{
  "Comment":"Sample Step Function",
  "StartAt":"First Glue Job",
  "States": {
      "First Glue Job":{
         "Type": "Task",
         "Resource": "arn:aws:states:::glue:startJobRun.sync",
         "Parameters": {
            "JobName": "GlueJob-Firstjob"
      },
      "Next": "Second Glue Job"
    },
    "Second Glue Job":{
         "Type": "Task",
         "Resource": "arn:aws:states:::glue:startJobRun.sync",
         "Parameters": {
            "JobName": "GlueJob-Secondjob"
      },
      "Next": "Third Glue Job"
    },
    "Third Glue Job":{
         "Type": "Task",
         "Resource": "arn:aws:states:::glue:startJobRun.sync",
         "Parameters": {
            "JobName": "GlueJob-Thirdjob"
      },
      "End": true …

我在AWS Step Functions 中定义了一个管道。一个步骤被定义为Fargate Task，它拉取一个 docker 镜像并运行一些 python 代码。我惊讶地发现，如果在 Fargate 任务中运行的容器遇到运行时错误，Step Functions 不会捕获失败的任务并照常继续管道（将 Fargate 任务设置为成功），但根据文档，管道一旦发生这种情况，应该会失败。

这是阶跃函数定义：

{
  "Comment": "My state machine",
  "StartAt": "MyFargateTask",
  "States": {
    "MyFargateTask": {
      "Type": "Task",
      "Resource": "arn:aws:states:::ecs:runTask.sync",
      "InputPath": "$",
      "Parameters": {
        "Cluster": "my-cluster",
        "TaskDefinition": "arn:aws:ecs:us-east-1:617090640476:task-definition/my-task:1",
        "LaunchType": "FARGATE",
        "NetworkConfiguration": {
          "AwsvpcConfiguration": {
            "Subnets": [
              "subnet-xxxxxxxxxxxxxxxxx",
              "subnet-yyyyyyyyyyyyyyyyy"
            ],
            "AssignPublicIp": "ENABLED"
          }
        },
      },
      "Next": "Done"
    },
    "Done": {
      "Type": "Succeed"
    }
  }
}

我已经为 Fargate 容器尝试了以下简单的 python 代码：

def main():
    raise Exception("foobar") …