如何在.net核心中为多个请求运行api

Question

我有一个场景，我尝试从 UI 上传文件，然后在 .net 核心中我创建了一个端点，将文件推送到 azure blob 存储中，之后我编写了一个自动触发并生成一个 azure 函数调用另一个 API，该 API 进一步处理该文件并编写某种 SQL 查询，并在处理文件后生成一些数据，然后我再次将该生成的文件推送到 blob 到某个不同的容器。但是我面临一个问题，如果我上传单个文件，它对我来说很好，但是当我上传一个文件然后我上传另一个文件时，我的第二个文件丢弃了我的第一个数据处理并且在我的第二个容器中我只得到一个导出的文件用于第二个上传的文件。所以我希望尽可能多的用户可以上传文件，但所有进程都应该运行它不应该丢弃任何正在运行的进程。 如果你们建议我任何对我有帮助的方法。如果需要，我会分享我的代码

上传文件端点

  [HttpPost]
  [Route("uploadFileToBlob")]
  public async Task<IActionResult> UploadFileToBlobStorage(IFormFile file)
    {
            var userId = HttpContext.User.Identity.Name;
            var accessToken = Request.Headers[HeaderNames.Authorization];
            _manageClaim.UploadExcelToBlob(file, accessToken, userId);
            return Ok(new { Message = "Your file is uploaded and processing..."});
    }

此端点的存储库

 public void UploadExcelToBlob(IFormFile file, string accessToken, string userId)
        {

            CloudStorageAccount storageAccount = CloudStorageAccount.Parse(blobConnectionString);

            // Create the blob client.
            CloudBlobClient blobClient = storageAccount.CreateCloudBlobClient();

            // Retrieve reference to a previously created container.
            CloudBlobContainer container = blobClient.GetContainerReference("infolder");

            // Retrieve reference to a blob named "test.xlsx"
            CloudBlockBlob blockBlobReference = container.GetBlockBlobReference(file.FileName);

            container.Metadata.Clear();
            container.Metadata.Add("bearerToken", accessToken);
            container.SetMetadata();

            Stream myBlob = file.OpenReadStream();
            blockBlobReference.UploadFromStream(myBlob);

            var currentUser = _automationUsers.GetUserById(int.Parse(userId));

            BatchClaimStatus batchClaim = new BatchClaimStatus()
            {
                UserId = int.Parse(userId),
                UserName = currentUser.FirstName + ' ' + currentUser.LastName,
                InputFileName = file.FileName,
                IsExported = false,
                ErrorMessage = null,
                OutputFileName = null,
                UpdatedDate = DateTime.Now,
                UploadedDate = DateTime.Now
            };
            _blobRepository.AddImportedFileToDb(batchClaim);
        }

API 端点 当我从 UI 上传任何文件时，此端点会自动命中

[HttpPost]
[Route("getBatchClaimData")]
public async Task<IActionResult> ExportToCsv(List<CsvPropModel> model)
{
await _manageClaim.GetClaimData(model);
return Ok();
}

这是上述端点的存储库

 public async Task<bool> GetClaimData(List<CsvPropModel> csvProps)
        {

            var dataRange = new List<ClaimsResponse>();
            ClaimsResponse claims = new ClaimsResponse();

            foreach (var item in csvProps)
            {
                InvoiceResult invoice = _manage.GetInvoiceDetailsByInvoiceId(item.InvoiceId, item.NickName);
                EligiblePaymentStatusParameters eligibleParams = Converter.ConvertInvoiceToEligibleParams(invoice);

                claims = await GetClaimDetails(eligibleParams, invoice);

                claims.InvKey = invoice.InvKey;
                claims.NickName = invoice.NickName;
                claims.InvoiceNumber = invoice.InvNbrDisplay;

                dataRange.Add(claims);
            }

            var csvRecord = ClaimDataHelper.ConvertToCsvReCord(dataRange);
            var fileName = @$"{ DateTime.Now.ToShortDateString() }_ExportedBatchClaim_{ DateTime.Now.ToShortTimeString() }.csv";
            MemoryStream ms = new MemoryStream();
            using (var memoryStream = new MemoryStream())
            {
                using (var writer = new StreamWriter(memoryStream))
                {
                    using (var csv = new CsvWriter(writer, CultureInfo.InvariantCulture))
                    {
                        csv.WriteRecords(csvRecord);
                        CloudStorageAccount storageAccount = CloudStorageAccount.Parse(blobConnectionString);

                        // Create the blob client.
                        CloudBlobClient blobClient = storageAccount.CreateCloudBlobClient();

                        // Retrieve reference to a previously created container.
                        CloudBlobContainer container = blobClient.GetContainerReference("outfolder");

                        // Retrieve reference to a blob named "test.xlsx"
                        CloudBlockBlob blockBlobReference = container.GetBlockBlobReference(fileName);
                        memoryStream.Position = 0;
                        memoryStream.CopyTo(ms);
                        ms.Position = 0;
                        await blockBlobReference.UploadFromStreamAsync(ms);
                        ms.Flush();

                        
                        var importedData = _blobRepository.GetLastImportedData();
                        
                        BatchClaimStatus batchClaim = new BatchClaimStatus()
                        {
                            ErrorMessage = null,
                            InputFileName = importedData.InputFileName,
                            IsExported = true,
                            OutputFileName = fileName,
                            UpdatedDate = DateTime.Now,
                            UploadedDate = importedData.UploadedDate,
                            UserId = importedData.UserId,
                            UserName = importedData.UserName
                        };

                        _blobRepository.UpdateExportedFileToDb(batchClaim, importedData.Id);
                        
                    }
                }
            }
            return true;
        }

Answer 1

我在您的上述代码中没有看到任何可能导致这种情况的问题，尽管您应该始终尝试进行任何 IO 绑定操作，例如_blobRepository.UpdateExportedFileToDb(batchClaim, importedData.Id) 或 _manageClaim.UploadExcelToBlob(file, accessToken, userId) async 以提高并发性。

实际问题似乎与 Azure Function blob 触发器有关。有时您可能会在 blob 触发器（特别是函数消费计划）中遇到问题，例如空闲、缺少触发器或队列处理延迟。

对于高吞吐场景，速度和可靠性，如果您使用存储V2，建议使用EventGrid trigger作为Function。您可以参考this 了解更多详情和此example。