AWS Serverless Workshop - build serverless application with AWS Lambda

只言片语

s3 是 serverless 的；
sam 是在 CloudFormation 的基础上的简化；
sam 用于在本地模拟 lambda 运行环境
lambda 背后是用 container 来运行的；
lambda 是一种计算资源；
lambda 是 stateless 的，若要持久化存储，则需要使用外部存储；
CloudFormation 等价于 terraform ；
不需要关心 Scalability、Utilization、服务器管理；
在 vpc 内创建 lambda 的主要目的在于，要访问内部的一些资源（不建议使用，因此在 code start 过程中多了两个步骤）

AWS Lambda 实验配置

名称	配置
User name	serverless-workshop
Password	8#@ATn[BWdp)+p-)i}lptfo)34$wVPO4&][RTarJ
Access key ID	AKIAO7SWN73EOHZRA5YQ
Secret access key	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
Console login link	https://330252173168.signin.amazonaws.cn/console
s3 bucket name	serverless-workshop
stack	workshop-fei

sam 使用

本地运行（sam local commands）

对应 https://github.com/lazydragon/asap/tree/master/sam

echo example

write a template.yaml config of your serverless application
please check here for property details: https://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/aws-resource-lambda-function.html
sync invoke a lambda called HelloWorld

➜  sam git:(master) echo '{}' | sam local invoke "HelloWorld"
2018/04/10 11:29:35 Successfully parsed template.yaml
2018/04/10 11:29:35 Connected to Docker 1.37
2018/04/10 11:29:35 Runtime image missing, will pull....
2018/04/10 11:29:35 Fetching lambci/lambda:python3.6 image for python3.6 runtime...
python3.6: Pulling from lambci/lambda
5be106c3813f: Pull complete
e240967675e1: Pull complete
e1dddd3e665c: Pull complete
5898d3e42308: Pull complete
b61aa9f36223: Pull complete
Digest: sha256:02f773a97b4e1663e828d97a7f14b8f661453bbbe829e24d64618870aa662e74
Status: Downloaded newer image for lambci/lambda:python3.6
2018/04/10 11:32:02 Reading invoke payload from stdin (you can also pass it from file with --event)
2018/04/10 11:32:02 Invoking sync.lambda_handler (python3.6)
2018/04/10 11:32:07 WARNING: No AWS credentials found. Missing credentials may lead to slow startup times as detailed in https://github.com/awslabs/aws-sam-local/issues/134
2018/04/10 11:32:07 Mounting /Users/sunfei/workspace/Sandbox/asap/sam/python as /var/task:ro inside runtime container
START RequestId: 631fbbe8-985d-422f-acdd-5a4b125df622 Version: $LATEST
END RequestId: 631fbbe8-985d-422f-acdd-5a4b125df622
REPORT RequestId: 631fbbe8-985d-422f-acdd-5a4b125df622 Duration: 9 ms Billed Duration: 100 ms Memory Size: 128 MB Max Memory Used: 19 MB

"Hello World"


➜  sam git:(master)

go example

build go code

go get github.com/aws/aws-lambda-go/lambda
GOOS=linux GOARCH=amd64 go build -o main main.go

invoke lambda

➜  asap git:(master) go get github.com/aws/aws-lambda-go/lambda
➜  asap git:(master) cd sam/go/
➜  go git:(master) GOOS=linux GOARCH=amd64 go build -o main main.go
➜  go git:(master) ✗ cd ..
➜  sam git:(master) ✗ echo '{}' | sam local invoke "HelloWorldGo"
2018/04/10 13:07:42 Successfully parsed template.yaml
2018/04/10 13:07:42 Connected to Docker 1.37
2018/04/10 13:07:42 Runtime image missing, will pull....
2018/04/10 13:07:42 Fetching lambci/lambda:go1.x image for go1.x runtime...
go1.x: Pulling from lambci/lambda
5be106c3813f: Already exists
e240967675e1: Already exists
d537dac75d99: Pull complete
5e61f8bb93d6: Pull complete
Digest: sha256:0c3d4ad28d7226c5b916a04051173ba3c21e2b7d213d63281d67e3aeac4a901b
Status: Downloaded newer image for lambci/lambda:go1.x
2018/04/10 13:07:46 Reading invoke payload from stdin (you can also pass it from file with --event)
2018/04/10 13:07:46 Invoking main (go1.x)
2018/04/10 13:07:51 WARNING: No AWS credentials found. Missing credentials may lead to slow startup times as detailed in https://github.com/awslabs/aws-sam-local/issues/134
2018/04/10 13:07:51 Mounting /Users/sunfei/workspace/Sandbox/asap/sam/go as /var/task:ro inside runtime container
START RequestId: 46be9759-cb7f-1738-bd43-0389e751fa27 Version: $LATEST
END RequestId: 46be9759-cb7f-1738-bd43-0389e751fa27
REPORT RequestId: 46be9759-cb7f-1738-bd43-0389e751fa27	Duration: 1.40 ms	Billed Duration: 100 ms	Memory Size: 128 MB	Max Memory Used: 5 MB
"Hello ƛ!"


➜  sam git:(master) ✗

部署到远端（sam deployment commands）

需要创建 /Users/sunfei/.aws 目录，并将 config 配置放在下面，config 内容如下

[default]
aws_access_key_id = AKIAO7SWN73EOHZRA5YQ
aws_secret_access_key = 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
region = cn-north-1

之后就可以

package and upload source code to s3

➜  sam git:(master) ✗ sam package --template-file template.yaml --s3-bucket serverless-workshop --output-template-file output.yaml
Uploading to 1df966252fa98952589fd07f39e865f3  175 / 175.0  (100.00%)100.00%)
Successfully packaged artifacts and wrote output template to file output.yaml.
Execute the following command to deploy the packaged template
aws cloudformation deploy --template-file /Users/sunfei/workspace/Sandbox/asap/sam/output.yaml --stack-name <YOUR STACK NAME>
➜  sam git:(master) ✗

deploy the serverless application

会上传

➜  sam git:(master) ✗ sam deploy --template-file output.yaml --stack-name workshop-fei --capabilities CAPABILITY_IAM

Waiting for changeset to be created..
Waiting for stack create/update to complete
Successfully created/updated stack - workshop-fei
➜  sam git:(master) ✗

describe output of the function

➜  sam git:(master) ✗ aws cloudformation describe-stacks --stack-name workshop-fei
{
    "Stacks": [
        {
            "StackId": "arn:aws-cn:cloudformation:cn-north-1:330252173168:stack/workshop-fei/aea928a0-3c72-11e8-b953-50fa18a2d236",
            "Description": "serverless hello world",
            "Tags": [],
            "Outputs": [
                {
                    "Description": "Go Lambda function name generated",
                    "OutputKey": "GoLambdaName",
                    "OutputValue": "workshop-fei-HelloWorldGo-ILLS7Y2H41EP"
                },
                {
                    "Description": "Lambda function name generated",
                    "OutputKey": "LambdaName",
                    "OutputValue": "workshop-fei-HelloWorld-SZSGC6WWWHDX"
                }
            ],
            "EnableTerminationProtection": false,
            "CreationTime": "2018-04-10T03:53:08.927Z",
            "Capabilities": [
                "CAPABILITY_IAM"
            ],
            "StackName": "workshop-fei",
            "NotificationARNs": [],
            "StackStatus": "CREATE_COMPLETE",
            "DisableRollback": false,
            "RollbackConfiguration": {},
            "ChangeSetId": "arn:aws-cn:cloudformation:cn-north-1:330252173168:changeSet/awscli-cloudformation-package-deploy-1523332388/c16c43e7-0a47-4ce5-baba-e9fe95a501f9",
            "LastUpdatedTime": "2018-04-10T03:53:14.414Z"
        }
    ]
}
➜  sam git:(master) ✗

test invoke of lambda

对应远程 invoke

➜  sam git:(master) ✗ aws lambda invoke --function-name workshop-fei-HelloWorld-SZSGC6WWWHDX workshop-fei-HelloWorld-SZSGC6WWWHDX.output
{
    "ExecutedVersion": "$LATEST",
    "StatusCode": 200
}
➜  sam git:(master) ✗

➜  sam git:(master) ✗ aws lambda invoke --function-name workshop-fei-HelloWorldGo-ILLS7Y2H41EP workshop-fei-HelloWorldGo-ILLS7Y2H41EP.output
{
    "ExecutedVersion": "$LATEST",
    "StatusCode": 200
}
➜  sam git:(master) ✗

delete stack

aws cloudformation delete-stack --stack-name <YOUR STACK NAME>

code start 测试

对应 https://github.com/lazydragon/asap/tree/master/code_start

create s3 bucket and deploy

➜  code_start git:(master) ✗ ./deploy.sh serverless-workshop workshop-fei
Uploading to 0cad4709cde641840f0d371bbeb20ea1  374 / 374.0  (100.00%)
Successfully packaged artifacts and wrote output template to file output.yaml.
Execute the following command to deploy the packaged template
aws cloudformation deploy --template-file /Users/sunfei/workspace/Sandbox/asap/code_start/output.yaml --stack-name <YOUR STACK NAME>

Waiting for changeset to be created..
Waiting for stack create/update to complete
Successfully created/updated stack - workshop-fei
{
    "Stacks": [
        {
            "StackId": "arn:aws-cn:cloudformation:cn-north-1:330252173168:stack/workshop-fei/aea928a0-3c72-11e8-b953-50fa18a2d236",
            "Description": "A function to experiment cold start time when code starts",
            "Tags": [],
            "Outputs": [
                {
                    "Description": "Lambda function name generated",
                    "OutputKey": "LambdaName",
                    "OutputValue": "workshop-fei-CodeStart-BNP2UW5HUAXB"
                }
            ],
            "EnableTerminationProtection": false,
            "CreationTime": "2018-04-10T03:53:08.927Z",
            "Capabilities": [
                "CAPABILITY_IAM"
            ],
            "StackName": "workshop-fei",
            "NotificationARNs": [],
            "StackStatus": "UPDATE_COMPLETE",
            "DisableRollback": false,
            "RollbackConfiguration": {},
            "ChangeSetId": "arn:aws-cn:cloudformation:cn-north-1:330252173168:changeSet/awscli-cloudformation-package-deploy-1523337864/32c0a2fe-07ca-4a6a-969a-93506e36fc33",
            "LastUpdatedTime": "2018-04-10T05:24:30.538Z"
        }
    ]
}
➜  code_start git:(master) ✗

测试

代码未调整前

➜  code_start git:(master) ✗ python invoke.py -mw 1 -it 30 -l workshop-fei-CodeStart-BNP2UW5HUAXB
BHSTMG: This is the 25 time this container is used

BHSTMG: This is the 26 time this container is used

BHSTMG: This is the 27 time this container is used

BHSTMG: This is the 28 time this container is used

BHSTMG: This is the 29 time this container is used

BHSTMG: This is the 30 time this container is used

3UFMV7: This is the 16 time this container is used

3UFMV7: This is the 17 time this container is used

3UFMV7: This is the 18 time this container is used

3UFMV7: This is the 19 time this container is used

3UFMV7: This is the 20 time this container is used

3UFMV7: This is the 21 time this container is used

3UFMV7: This is the 22 time this container is used

BHSTMG: This is the 31 time this container is used

BHSTMG: This is the 32 time this container is used

BHSTMG: This is the 33 time this container is used

BHSTMG: This is the 34 time this container is used

BHSTMG: This is the 35 time this container is used

BHSTMG: This is the 36 time this container is used

BHSTMG: This is the 37 time this container is used

BHSTMG: This is the 38 time this container is used

BHSTMG: This is the 39 time this container is used

BHSTMG: This is the 40 time this container is used

3UFMV7: This is the 23 time this container is used

3UFMV7: This is the 24 time this container is used

3UFMV7: This is the 25 time this container is used

3UFMV7: This is the 26 time this container is used

BHSTMG: This is the 41 time this container is used

BHSTMG: This is the 42 time this container is used

BHSTMG: This is the 43 time this container is used

➜  code_start git:(master) ✗
➜  code_start git:(master) ✗ python invoke.py -mw 10 -it 30 -l workshop-fei-CodeStart-BNP2UW5HUAXB | awk '{print $1}' | sort | uniq -c
  30
   5 3UFMV7:
   1 A5I0MD:
   1 ACD1AC:
   9 BHSTMG:
   1 D0DBUB:
   1 I7L44J:
   1 ST0DQE:
   1 T49RQV:
   1 VB79IM:
   1 XEVO3B:
   8 YXGXVZ:
➜  code_start git:(master) ✗ python invoke.py -mw 10 -it 30 -l workshop-fei-CodeStart-BNP2UW5HUAXB | awk '{print $1}' | sort | uniq -c
  30
   3 A5I0MD:
   4 ACD1AC:
   2 BHSTMG:
   3 D0DBUB:
   4 I7L44J:
   4 ST0DQE:
   4 T49RQV:
   2 VB79IM:
   3 XEVO3B:
   1 YXGXVZ:
➜  code_start git:(master) ✗
➜  code_start git:(master) ✗ python invoke.py -mw 10 -it 100 -l workshop-fei-CodeStart-BNP2UW5HUAXB | awk '{print $1}' | sort | uniq -c
 100
  10 A5I0MD:
  12 ACD1AC:
  11 BHSTMG:
  11 D0DBUB:
   2 I7L44J:
   9 ST0DQE:
  10 T49RQV:
  12 VB79IM:
  12 XEVO3B:
  11 YXGXVZ:
➜  code_start git:(master) ✗ python invoke.py -mw 10 -it 100 -l workshop-fei-CodeStart-BNP2UW5HUAXB | awk '{print $1}' | sort | uniq -c
 100
   1 28EK2L:
   7 3UFMV7:
  10 A5I0MD:
   9 ACD1AC:
  10 BHSTMG:
   9 D0DBUB:
   5 I7L44J:
   9 ST0DQE:
   9 T49RQV:
  11 VB79IM:
  10 XEVO3B:
  10 YXGXVZ:
➜  code_start git:(master) ✗

调整代码后（只需要调整 src/lambda.py）

 18 def lambda_handler(event, context):
 19     global count, container_id
 20     # count add 1
 21     count += 1
 22     time.sleep(1)   # 增加
 23
 24     return (container_id, count)

➜  code_start git:(master) ✗ python invoke.py -mw 10 -it 10 -l workshop-fei-CodeStart-BNP2UW5HUAXB | awk '{print $1}' | sort | uniq -c
  10
   1 6YHLKQ:
   1 7Z3BLS:
   1 9WSNU1:
   1 E9YQX9:
   1 MZJAW1:
   1 OPQQH7:
   1 RVXGOU:
   1 UH5YL7:
   1 UL920R:
   1 WE4UXH:
➜  code_start git:(master) ✗ python invoke.py -mw 10 -it 100 -l workshop-fei-CodeStart-BNP2UW5HUAXB | awk '{print $1}' | sort | uniq -c
 100
  10 6YHLKQ:
  10 7Z3BLS:
  10 9WSNU1:
  10 E9YQX9:
  10 MZJAW1:
  10 OPQQH7:
  10 RVXGOU:
  10 UH5YL7:
  10 UL920R:
  10 WE4UXH:
➜  code_start git:(master) ✗
➜  code_start git:(master) ✗ python invoke.py -mw 10 -it 30 -l workshop-fei-CodeStart-BNP2UW5HUAXB | awk '{print $1}' | sort | uniq -c
  30
   3 1L15GC:
   3 4C4USH:
   3 5JLHWK:
   3 90Q7GJ:
   3 K1U0HS:
   3 MZJAW1:
   3 QHZG2L:
   3 T0FP54:
   3 TCU5QA:
   3 WE4UXH:
➜  code_start git:(master) ✗

结论：

在未进行代码调整前，worker 数量和 container 的数量是不一致的，每个 container 被复用的次数也是不同的；
在为 lambda.py 函数增加延迟后，可以看到 worker 数量和 container 的数量是相同的，并且每个 container 被复用的次数也是相同的；

performance 测试

对应 https://github.com/lazydragon/asap/tree/master/performance

略

结论：

资源使用主要基于 memory 进行控制（CPU 的使用和 memory 成线性关系，但不能直接设置）
在 memory 设置为 2048MB 时，程序执行实现最少，性能最大化；

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

AWS Serverless Workshop - lambda.md

AWS Serverless Workshop - lambda.md

AWS Serverless Workshop - build serverless application with AWS Lambda

只言片语

AWS Lambda 实验配置

sam 使用

本地运行（sam local commands）

echo example

go example

部署到远端（sam deployment commands）

code start 测试

performance 测试

Files

AWS Serverless Workshop - lambda.md

Latest commit

History

AWS Serverless Workshop - lambda.md

File metadata and controls

AWS Serverless Workshop - build serverless application with AWS Lambda

只言片语

AWS Lambda 实验配置

sam 使用

本地运行（sam local commands）

echo example

go example

部署到远端（sam deployment commands）

code start 测试

performance 测试