Azure app: how to log POSTed JSON to my app? - azure

I went to my App services > my app > Monitoring > Diagnostic logs
And there I enabled both: Application Logging (Blob) and Web server logging (Storage)
The above settings started logging a .log file in which every row is an HTTP request to my app, with these headers:
#Fields: date time s-sitename cs-method cs-uri-stem cs-uri-query s-port cs-username c-ip cs(User-Agent) cs(Cookie) cs(Referer) cs-host sc-status sc-substatus sc-win32-status sc-bytes cs-bytes time-taken
I'm interested in logging the POSTed JSON for the requests that the server is responding HTTP 500. How can I achieve this from the Azure portal?

How can I achieve this from the Azure portal?
Azure Web App doesn't provide this feature default. I suggest you add a event handler for application level exception. In this handler, you could log the request body or the information of exception. Code below is for your reference.
protected void Application_Error(object sender, EventArgs e)
{
string req_JSON;
using (StreamReader reader = new StreamReader(Request.InputStream))
{
req_JSON = reader.ReadToEnd();
}
//you could log the json to anywhere you want
Exception excpt = Server.GetLastError().GetBaseException();
//you also could log the exception which encounted at server
}

Not clear what you are using to write your app, but if you are using .net you would need to use the System.Diagnostics.Trace namespace.
in this case since you are getting a 5XX http error, I asume there is some sort of exception being thrown that is missing a try/catch.
Once you find that, then you add something like:
System.Diagnostics.Trace.TraceError("If you're seeing this, something bad happened " + postedJSON);
This will result in the json being added to the application log.
You can get more info about App Service logging here: https://learn.microsoft.com/en-us/azure/app-service/web-sites-enable-diagnostic-log

Related

Azure App Service Hijacks Response for HTTP STATUS 500 Errors

I have an app service that is working as expected in that I can get to pages, log in, and perform a search.
In the event of errors I have the below code in place in the MVC Controller (not an API endpoint)
{
_log.Error("Customer Search Failed", ex);
Response.StatusCode = 500;
return new JsonResult()
{
Data = new { Success = false, Error = "An error occured.", Amount = 0 },
JsonRequestBehavior = JsonRequestBehavior.AllowGet,
ContentEncoding = System.Text.Encoding.UTF8
};
}
Locally, I get the following response along with an HTTP Status Code of 500:
{"Success":false,"Error":"An error occured.","Amount":0}
In Azure, though, I get this response:
The page cannot be displayed because an internal server error has occurred.
I have written a test app to reproduce this in my own Azure environment so that there wouldn't be anything company Azure Resources that could cause this and I got the same behavior. No application gateways, VNets, anything like that. The app service itself is effectively seeing the 500 HTTP Status Code and then sending its own response vs what I want to send.
Does anyone know why this is happening and/or how to prevent this from happening?
The problem was that Azure must set up standard responses for non-200 HTTP Status Codes. This overrode what my Controllers were returning.
I had to explicitly remove the 500 httpHeader via the web.config in order to see the JSON response I was expecting.
<httpErrors errorMode="Detailed">
<remove statusCode="500" />
</httpErrors>

No exceptions or stack traces in Azure Application Insights

I have an ASP.NET Core 3.1 solution deployed into an Azure Web App hooked up to Application Insights. I can't for the life of me get exceptions and stack traces to log into Application Insights, instead I get a basic request trace with no exception information attached:
I've tried most combinations of setting up logging/application insights telemetry, here are some of the things I've tried:
services.AddApplicationInsightsTelemetry(); in the ConfigureServices() method of Startup.cs
Adding logging.AddApplicationInsights(); to my logging builder in Program.cs
Removing the custom error page exception handler in case that was affecting things
I have the APPINSIGHTS_INSTRUMENTATIONKEY environment variable set on my Web App in Azure.
I'm using the following code to generate exceptions in Application Insights:
[AllowAnonymous]
[Route("autoupdate")]
public async Task<IActionResult> ProfileWebhook()
{
var formData = await this.Request.ReadFormAsync();
var config = TelemetryConfiguration.CreateDefault();
var client = new TelemetryClient(config);
client.TrackException(new Exception(string.Join("~", formData.Keys)));
logger.LogError(new Exception(string.Join("~", formData.Keys)), "Fail");
throw new Exception(string.Join("~", formData.Keys));
}
Nothing is working and I'm going crazy! Any help greatly appreciated.
Usually, Application insights will guarantee that all the kinds of telemetries(like exceptions, trace, event etc.) will be arrived around 5 minutes, please refer this doc: How long does it take for telemetry to be collected?. But there is still a chance that it will take a longer time due to beckend issue(a very small chance).
If you're using visual studio, you can check if the telemetry is sent or not via Application Insights search.
You can also check if you're using a correct IKey, or if you have enabled sampling.
But if it keeps this behavior in your side, you should consider contacting MS support to find the root cause.
Hope it helps.

Need help investigating Azure App Service .NET Core 2.1 ArgumentOutOfRangeException and System.OverflowException

Today I've adjusted the App Service Log settings for my Azure App Serivce: a .NET Core 2.1 Web API. Soon after I changed the Application Logging (file system) setting from Information to Warning, my application stopped responding. I immediately kicked of a .NET Trace using the Diagnose and solve problems feature.
This Trace report showed the following:
system.private.corelib!
microsoft.aspnetcore.hosting!
System.Private.CoreLib!System.Runtime.CompilerServices.AsyncMethodBuilderCore.Start
microsoft.aspnetcore.hosting.il!Microsoft.AspNetCore.Hosting.Views.ErrorPage.ExecuteAsync
microsoft.aspnetcore.hosting!
System.Private.CoreLib!System.Runtime.CompilerServices.AsyncMethodBuilderCore.Start
microsoft.aspnetcore.hosting.il!Microsoft.Extensions.RazorViews.BaseView.ExecuteAsync
microsoft.aspnetcore.hosting!
microsoft.aspnetcore.hosting!
System.Private.CoreLib!System.Runtime.CompilerServices.AsyncTaskMethodBuilder`1+AsyncStateMachineBox`1+<>c[System.Threading.Tasks.VoidTaskResult,Microsoft.AspNetCore.Server.Kestrel.Core.Internal.Http.HttpProtocol+<ProcessRequests>d__188`1[Microsoft.AspNetCore.Hosting.Internal.HostingApplication+Context]].<.cctor>b__9_0
system.private.corelib!
System.Private.CoreLib!System.Runtime.CompilerServices.AsyncTaskMethodBuilder`1+AsyncStateMachineBox`1[System.Threading.Tasks.VoidTaskResult,Microsoft.AspNetCore.Server.Kestrel.Core.Internal.Http.HttpProtocol+<ProcessRequests>d__188`1[Microsoft.AspNetCore.Hosting.Internal.HostingApplication+Context]].MoveNext
system.private.corelib!
system.private.corelib!
system.private.corelib!
system.private.corelib!
system.private.corelib!
Same time, the Azure Blob log showed these lines:
2019-07-06 14:04:06.476 +00:00 [Error] Microsoft.AspNetCore.Server.Kestrel: Connection id "0HLO233NHLK6Q", Request id "0HLO233NHLK6Q:000003BF": An unhandled exception was thrown by the application.
System.OverflowException: Arithmetic operation resulted in an overflow.
at System.IO.StreamWriter.Write(String value)
at Microsoft.Extensions.RazorViews.BaseView.Write(String value)
at Microsoft.AspNetCore.Hosting.Views.ErrorPage.ExecuteAsync()
at Microsoft.Extensions.RazorViews.BaseView.ExecuteAsync(HttpContext context)
at Microsoft.AspNetCore.Server.Kestrel.Core.Internal.Http.HttpProtocol.ProcessRequests[TContext](IHttpApplication`1 application)
2019-07-06 14:04:19.245 +00:00 [Error] Microsoft.AspNetCore.Server.Kestrel: Connection id "0HLO233NHLK6T", Request id "0HLO233NHLK6T:000000EF": An unhandled exception was thrown by the application.
System.ArgumentOutOfRangeException: Index and count must refer to a location within the buffer.
Parameter name: chars
at System.Text.EncoderNLS.GetBytes(Char[] chars, Int32 charIndex, Int32 charCount, Byte[] bytes, Int32 byteIndex, Boolean flush)
at System.IO.StreamWriter.Flush(Boolean flushStream, Boolean flushEncoder)
at System.IO.StreamWriter.Write(String value)
at Microsoft.AspNetCore.Hosting.Views.ErrorPage.ExecuteAsync()
at Microsoft.Extensions.RazorViews.BaseView.ExecuteAsync(HttpContext context)
at Microsoft.AspNetCore.Server.Kestrel.Core.Internal.Http.HttpProtocol.ProcessRequests[TContext](IHttpApplication`1 application)
These stacktraces don't show any of my own code, they seem framework related.
Is there a message or a hint in these error stack traces that I can use to prevent these situations?
Many thanks in advance
I know it's too late but maybe my answer will help someone else.
I received the same exceptions as you. But it was not Azure App Service but IIS. The problem was in incorrect configuration somewhere in Startup.cs. You webhost tried to show error page but at the same time App Service (or IIS in my case) tried to shut down or restart it.
This thread helped me a lot https://github.com/aspnet/KestrelHttpServer/issues/1737

Azure backend return 500 in PATCH operation

I am desperately trying to debug an error 500 only when I try to update an object from my xamarin.Forms offline DB to Azure. I am using Azure Mobile Client.
I set all the logging to ON in azure, then I downloaded the log. I can see the generic error, but nothing useful.
<failedRequest url="https://MASKED:80/tables/Appel/9A3342A2-0598-4126-B0F6-2999B524B4AE"
siteId="Masked"
appPoolId="Masked"
processId="6096"
verb="PATCH"
remoteUserName=""
userName=""
tokenUserName="IIS APPPOOL\Masked"
authenticationType="anonymous"
activityId="{80000063-0000-EA00-B63F-84710C7967BB}"
failureReason="STATUS_CODE"
statusCode="500"
triggerStatusCode="500"
timeTaken="625"
xmlns:freb="http://schemas.microsoft.com/win/2006/06/iis/freb"
>
The table that failed is the only one I extend with some virtual runtime calculated field of navigation field. But I add the [JsonIgnore] to stop AzureService to create field in the local DB (that work) or send it on the wire to the server. But I always got the 500 error, not exception when debugging the c# Azure backend too.
How I can find the stack trace or the "deep" reason for this 500 error in my backend?
For C# Mobile App backend, you could add the following code in the ConfigureMobileApp method of your Startup.MobileApp.cs file for including error details and return to your client side.
config.IncludeErrorDetailPolicy = IncludeErrorDetailPolicy.Always;
You could just capture the exception in your mobile application or leverage fiddler to capture the network traces when invoking the PATCH operation to retrieve the detailed error message.
Moreover, you are viewing the Failed Request Traces log, you need to check the Application logs. Details you could follow Enable diagnostics logging for web apps in Azure App Service.

How can I keep my Azure WebJob running without "Always On"

I have a continous webjob associated with a website and I am running that website on the Shared mode. I don't want to go to Always On option as there is no real need for my application. I only want to process the message when the calls are made to my website.
My issue is that the job keeps stopping after few minutes even though I am continuously calling a dummy keep alive method on my website at every 5 minute that posts a message to the queue that is monitored by that webjob.
My webjob is a simple console application built using the WebJob SDK that has a code like this
JobHost host = new JobHost(new JobHostConfiguration(storageConnictionSttring));
host.RunAndBlock();
and the message processing function looks like below:
public static void ProcessKeepAliveMessages([QueueTrigger("keepalive")] KeepAliveTrigger message)
{
Console.WriteLine("Keep Alive message called on :{0}", message.MessageTime);
}
The message log for the job basically says says
[03/05/2015 18:51:02 > 4660f6: SYS INFO] WebJob is stopping due to website shutting down
I don't mind if that happen this way, but when the website starts with the next call to keep alive, the webjob is not started. All the messages are queued till I go to the management dashboard or the SCM portal as shown below
https://mysite.scm.azurewebsites.net/api/continuouswebjobs
I can see the status like this:
[{"status":"Starting","detailed_status":"4660f6 - Starting\r\n","log_url":"https://mysite.scm.azurewebsites.net/vfs/data/jobs/continuous/WebJobs/job_log.txt","name":"WebJobs","run_command":"mysite.WebJobs.exe","url":"https://mysite.scm.azurewebsites.net/api/continuouswebjobs/WebJobs","extra_info_url":"https://mysite.scm.azurewebsites.net/azurejobs/#/jobs/continuous/WebJobs","type":"continuous","error":null,"using_sdk":true,"settings":{}}]
I would really appreciate if someone can help me understand what is going wrong here.
I've run into a similar problem. I have a website (shared mode) and an associated webjob (continuous type). Looking at webjob logs, I found that the job enters stopped state after about 15 min. of inactivity and stops reacting to trigger messages. It seems contradictory to the concept of continuous job concept but, apparently, to get it running truly continuously you have to subscribe to a paid website. You get what you pay for...
That said, my website needs to be used only about every few days and running in a shared mode makes perfect sense. I don't mind that the site needs a bit extra time to get started - as long as it restarts automatically. The problem with the webjob is that once stopped it won't restart by itself. So, my goal was to restart it with the website.
I have noticed that a mere look at the webjob from Azure Management Portal starts it. Following this line of thinking, I have found that fetching webjob properties is enough to switch it to the running state. The only trick is how to fetch the properties programmatically, so that restarting the website will also restart the webjob.
Because the call to fetch webjob properties must be authenticated, the first step is to go to Azure Management Portal and download the website publishing profile. In the publishing profile you can find the authentication credentials: username (usually $<website_name>) and userPWD (hash of the password). Copy them down.
Here is a function that will get webjob properties and wake it up (if not yet running):
class Program
{
static void Main(string[] args)
{
string websiteName = "<website_name>";
string webjobName = "<webjob_name>";
string userName = "<from_publishing_profile>";
string userPWD = "<from_publishing_profile>";
string webjobUrl = string.Format("https://{0}.scm.azurewebsites.net/api/continuouswebjobs/{1}", websiteName, webjobName);
var result = GetWebjobState(webjobUrl, userName, userPWD);
Console.WriteLine(result);
Console.ReadKey(true);
}
private static JObject GetWebjobState(string webjobUrl, string userName, string userPWD)
{
HttpClient client = new HttpClient();
string auth = "Basic " + Convert.ToBase64String(Encoding.UTF8.GetBytes(userName + ':' + userPWD));
client.DefaultRequestHeaders.Add("authorization", auth);
client.DefaultRequestHeaders.Accept.Add(new MediaTypeWithQualityHeaderValue("application/json"));
var data = client.GetStringAsync(webjobUrl).Result;
var result = JsonConvert.DeserializeObject(data) as JObject;
return result;
}
}
You can use a similar function to get all webjobs in your website (use endpoint https://<website_name>.scm.azurewebsites.net/api/webjobs). You may also look at the returned JObject to verify the actual state of the webjob and other properties.
If you want the WebJob to not stop you need to make sure your scm site is alive.
So the keep-alive requests should go to https://sitename.scm.azurewebsites.net and these requests need to be authenticated (basic auth using your deployment credentials).

Resources