将数据从 Azure Blob 存储复制到 AWS S3

Question

我是 Azure 数据工厂的新手，有一个有趣的需求。

我需要将文件从 Azure Blob 存储移动到 Amazon S3，最好使用 Azure 数据工厂。

但是 S3 不支持作为接收器；

https://docs.microsoft.com/en-us/azure/data-factory/copy-activity-overview

我还从我在这里阅读的各种评论中了解到，您不能直接从 Blob 存储复制到 S3 - 您需要在本地下载文件，然后将其上传到 S3。

有谁知道在数据工厂、SSIS 或 Azure Runbook 中可以执行此类操作的任何示例，我想一个选项是编写从数据工厂调用的 azure 逻辑应用程序或函数。

Answer 1

设法让一些东西在这方面工作 - 它可能对其他人有用。

我决定编写一个使用 HTTP 请求作为触发器的 azure 函数。

这两篇文章对我帮助很大；

如何在 Azure Functions 中使用 NuGet 包？

使用 C# 从 Azure Blob 复制到 AWS S3

如果您使用的是 Azure 函数 2.x，请注意我对 Nuget 包的回答。

这是代码 - 您可以根据需要修改其基础。我返回一个 JSON 序列化对象，因为 Azure 数据工厂需要它作为来自管道发送的 http 请求的响应；

#r "Microsoft.WindowsAzure.Storage"
#r "Newtonsoft.Json"
#r "System.Net.Http"

using System.Net;
using Microsoft.AspNetCore.Mvc;
using Microsoft.Extensions.Primitives;
using Newtonsoft.Json;
using Microsoft.WindowsAzure.Storage.Blob;
using System.Net.Http;
using Amazon.S3; 
using Amazon.S3.Model;
using Amazon.S3.Transfer;
using Amazon.S3.Util;


public static async  Task<IActionResult> Run(HttpRequest req, ILogger log)
{
    log.LogInformation("Example Function has recieved a HTTP Request");

    // get Params from query string
    string blobUri = req.Query["blobUri"];
    string bucketName = req.Query["bucketName"];

    // Validate query string
    if (String.IsNullOrEmpty(blobUri) || String.IsNullOrEmpty(bucketName)) {

        Result outcome = new Result("Invalid Parameters Passed to Function",false,"blobUri or bucketName is null or empty");
        return new BadRequestObjectResult(outcome.ConvertResultToJson());
    }

    // cast the blob to its type
    Uri blobAbsoluteUri = new Uri(blobUri);
    CloudBlockBlob blob = new CloudBlockBlob(blobAbsoluteUri);

    // Do the Copy
    bool resultBool = await CopyBlob(blob, bucketName, log);

    if (resultBool) { 
        Result outcome = new Result("Copy Completed",true,"Blob: " + blobUri + " Copied to Bucket: " + bucketName);
        return (ActionResult)new OkObjectResult(outcome.ConvertResultToJson());       
    }
    else {
        Result outcome = new Result("ERROR",false,"Copy was not successful Please review Application Logs");
        return new BadRequestObjectResult(outcome.ConvertResultToJson()); 
    }  
}

static async Task<bool> CopyBlob(CloudBlockBlob blob, string existingBucket, ILogger log) {

        var accessKey = "myAwsKey";
        var secretKey = "myAwsSecret";
        var keyName = blob.Name;

        // Make the client 
        AmazonS3Client myClient = new AmazonS3Client(accessKey, secretKey, Amazon.RegionEndpoint.EUWest1);

        // Check the Target Bucket Exists; 
        bool bucketExists = await AmazonS3Util.DoesS3BucketExistAsync (myClient,existingBucket);

        if (!bucketExists) {
            log.LogInformation("Bucket: " + existingBucket + " does not exist or is inaccessible to the application");
            return false;
        }

        // Set up the Transfer Utility
        TransferUtility fileTransferUtility = new TransferUtility(myClient);

        // Stream the file
        try {

            log.LogInformation("Starting Copy");

            using (var stream = await blob.OpenReadAsync()) {

                // Note: You need permissions to not be private on the source blob
                log.LogInformation("Streaming");

                await fileTransferUtility.UploadAsync(stream,existingBucket,keyName);

                log.LogInformation("Streaming Done");   
            }

            log.LogInformation("Copy completed");
        }
        catch (AmazonS3Exception e) {
                log.LogInformation("Error encountered on server. Message:'{0}' when writing an object", e.Message);
            }
        catch (Exception e) {
                log.LogInformation("Unknown encountered on server. Message:'{0}' when writing an object", e.Message);
                return false;
        }

        return true; 
    }

public class Result {

    public string result;
    public bool outcome;
    public string UTCtime;
    public string details; 

    public Result(string msg, bool outcomeBool, string fullMsg){
        result=msg;
        UTCtime=DateTime.Now.ToString("yyyy-MM-dd h:mm:ss tt");
        outcome=outcomeBool;
        details=fullMsg;
    }

    public string ConvertResultToJson() {
        return JsonConvert.SerializeObject(this);
    } 
}