Azure函数主机死于OutOfMemoryException，而不会触发 [英] Azure function host dies with OutOfMemoryException, without triggering
问题描述
快速版本:为什么我的函数宿主在5分钟后不执行任何操作会自行杀死它?
我有一个Azure函数，该函数使用nClam扫描blob文件中的病毒.它似乎可以正常工作，但是突然之间，即使在触发任何Blob之前，它也会自我杀死！它将在5分钟后关闭，并显示OutOfMemoryException:
I have an Azure function which uses nClam to scan blob files for viruses. It seems to work just fine, but suddenly it will kill it self even before triggering on any blob! It will just shutdown after 5 minutes with an OutOfMemoryException:
[18/9/2019 10:33:33] Host initialized (405ms)
[18/9/2019 10:33:33] Host started (812ms)
[18/9/2019 10:33:33] Job host started
Hosting environment: Production
Content root path: D:\git\TopoAPI\Antivirus\bin\Debug\netcoreapp2.1
Now listening on: http://0.0.0.0:7071
Application started. Press Ctrl+C to shut down.
[18/9/2019 10:33:38] Host lock lease acquired by instance ID '000000000000000000000000C913FBA0'.
[18/9/2019 10:38:46] An unhandled exception has occurred. Host is shutting down.
[18/9/2019 10:38:46] Microsoft.WindowsAzure.Storage: Exception of type 'System.OutOfMemoryException' was thrown. System.Private.CoreLib: Exception of type 'System.OutOfMemoryException' was thrown.
[18/9/2019 10:38:46] Stopping host...
[18/9/2019 10:38:46] Stopping JobHost
[18/9/2019 10:38:46] Job host stopped
[18/9/2019 10:38:46] Host shutdown completed.
Application is shutting down...
下面是我的扫描Blob触发函数:
Below is my scanning blob triggered function:
[FunctionName("scanImports")]
public static async Task Scan([BlobTrigger("imports/{newBlobName}", Connection = "BlobConnectionstring")]CloudBlockBlob newBlob, string newBlobName, ILogger log, ExecutionContext context)
{
var config = new ConfigurationBuilder().SetBasePath(context.FunctionAppDirectory).AddJsonFile("local.settings.json", optional: true, reloadOnChange: true).AddEnvironmentVariables().Build();
var clamClient = new ClamClient(config["ContainerAddress"], int.Parse(config["ContainerPort"]));
var blobFileStream = await newBlob.OpenReadAsync();
using (var memoryStream = new MemoryStream())
{
await blobFileStream.CopyToAsync(memoryStream);
var result = await clamClient.SendAndScanFileAsync(memoryStream.ToArray());
bool isClean = result.InfectedFiles == null || result.InfectedFiles.Count == 0;
// Check if newBlob is infected. If infected, move to quarantineBlob
if (!isClean)
{
CloudStorageAccount storageAccount = CloudStorageAccount.Parse(config["blobConnectionstring"]);
CloudBlobClient client;
CloudBlobContainer container;
CloudBlockBlob quarantineBlob;
client = storageAccount.CreateCloudBlobClient();
container = client.GetContainerReference(config["QuarantineBlobName"]);
await container.CreateIfNotExistsAsync();
quarantineBlob = container.GetBlockBlobReference(newBlobName);
quarantineBlob.Properties.ContentType = newBlob.Properties.ContentType;
await quarantineBlob.UploadFromStreamAsync(memoryStream);
await newBlob.DeleteAsync();
}
}
}
更新1:主机恰好在5分钟后死于OutOfMemoryException.我试图延长函数的超时时间，这没有什么区别. 在此期间，该进程将持续使用5-8％的cpu，而在死之前，该进程将使用1500 MB以上的内存.
Update 1: The host dies after exactly 5 minutes with a OutOfMemoryException. I have tried to extend the function timeout, that made no difference. During that time the process will constantly use 5-8% cpu and before it dies the process will be using over 1500 MB memory.
更新2:如果我从该函数中删除了所有代码，仅保留了一个log.info()语句，则主机在5分钟后仍会使用OutOfMemoryException自行杀死它
Update 2: If I remove all code from the function, leaving only a single log.info() statement, the host still kills it self after 5 minutes with an OutOfMemoryException
推荐答案
我的问题的原因是我试图触发的Blob中包含100.000+个文件. Microsoft将其定义为高级.
The cause of my problem was that the blob I was trying to trigger on had 100.000+ files in it. Microsoft defines that as High scale.
所以我的代码实际上没有问题.只是如果我尝试在高级" blob上触发，那么我的代码将什么也不会做，只会分配内存并消亡.
So my code actually had no problems. It is just that if I try to trigger on a "high scale" blob, then my code will do nothing but just allocate memory and die.
这篇关于Azure函数主机死于OutOfMemoryException，而不会触发的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！
