Hi All, I am trying to use the Serilog middleware to add more details to the logs but the diagnostic context is not logging it
public static IApplicationBuilder UseCustomSerilogRequestLogging(this IApplicationBuilder application) =>
application.UseSerilogRequestLogging(
options =>
{
options.EnrichDiagnosticContext = (diagnosticContext, httpContext) =>
{
var request = httpContext.Request;
var response = httpContext.Response;
var endpoint = httpContext.GetEndpoint();
diagnosticContext.Set("Host", request.Host);
diagnosticContext.Set("Protocol", request.Protocol);
diagnosticContext.Set("Scheme", request.Scheme);
if (request.QueryString.HasValue)
{
diagnosticContext.Set("QueryString", request.QueryString.Value);
}
if (endpoint is object)
{
diagnosticContext.Set("EndpointName", endpoint.DisplayName);
}
diagnosticContext.Set("ContentType", response.ContentType);
};
});
I dont see any differences in the logs adding this with the normal logging, so I am assuming is not logging any of the new stuff
Any one can help with above.. further details here https://stackoverflow.com/questions/62530177/how-to-apply-in-serilog-config-setting-to-generate-new-log-per-session-from-app
@maheshenator thanks for posting the SO link, will take a look :+1:
Hello, I am getting my feet wet with
I am trying to start from loosely based on the sample, but I am getting an exception,
Serilog first time, coming from a some time log4net background, so configuration, etc, is not entirely foreign to me.I am trying to start from loosely based on the sample, but I am getting an exception,
Inner Exception 1: ArgumentException: Configuration for Action<Serilog.Configuration.LoggerEnrichmentConfiguration> is not implemented.
any ideas why I might get that sort of exception, more or less out of the box/gate?
first thoughts, maybe I am missing an
.Enrich.... call. I suppose I could pear back to just .ReadFrom.Configuration(root) and see what that yields.
Quick question: I've written a wrapper around Serilog (to more easily capture source code information and to allow SMS logging). It supports dependency injection. My basic DI configuration (for Autofac) creates a singleton ILogger which is then wrapped by instances of the wrapper (IJ4JLogger).
Is that thread safe? Or should I use instance-based DI for both the wrapped ILogger and the wrapping IJ4JLogger? If I go instance-based is Serilog itself designed so that multiple ILogger's can, say, write to the same file sink? (That's why I went with a singleton ILogger in the first place)
Running into a weird challenge:
Public Sub LogClose(kronosId As String) Implements IAccessor.LogClose
Dim logEntry As New LogEntry With {
.KronosId = kronosId,
.FacIdentifier = _appConfig.FacIdentifier,
.Origin = _appConfig.Origin
}
Log.Debug("{Method}: {@LogEntry}", "LogClose", logEntry)
SetBaseRequest("Scads/Logs/ForceClose/{sessionId}/{gtnra}", Method.POST)
_request.RequestFormat = DataFormat.Json
_request.AddBody(logEntry)
_response = _client.Execute(_request)
End Subproduces this in the Seq:
Thoughts? logEntry is a simple class with simple properties...
Hello, and thank you in advance. I need to mitigate the risk of CRLF injection attacks in our application logging. (https://owasp.org/www-community/vulnerabilities/CRLF_Injection) . I reviewed the documentation but was unable to find guideance on this topic. We are using the Application Insights Sink (https://github.com/serilog/serilog-sinks-applicationinsights) it appears that a Custom ITelemetryConverter may be what we want but I wanted to confirm there isn't a solution already available or perhaps a better way to implement a solution.
Hi @KeithBarrows - this often happens when the properties are actually fields - not sure how they're declared in VB.NET, but if you post some more details of the
LogEntry class I might be able to spot it. HTH!
@nathanwinder Hi! Thanks for the link to the CRLF page on OWASP... Seems like a very weak vulnerability category?! (That page could just as well be named "Bug"!) :-D ... I think logs without CLRFs will be very odd indeed; imagine a multiline stack trace with no newlines... The good news is that all the logs sent to AI are JSON-encoded, so any injected CR or LF will be converted into the
\r or \n JSON escape sequence. AI will decode these to, for example, so line breaks in stack traces, but this is all main-line AI functionality and doesn't constitute any kind of vulnerability. HTH!
Public Class LogEntry
Public Id As Guid?
Public FacIdentifier As String
Public OriginId As String
Public Origin As String
Public DestinationId As String
Public Destination As String
Public KronosId As String
Public TimeIn As DateTime
Public TimeOut As DateTime?
End Class
If it's possible to change them to properties that's the most straightforward option; otherwise, there's a plug-in described in https://stackoverflow.com/questions/46152092/serilog-serializing-fields that can do the trick :-)
Hope this helps!
Been a little buried in the 2020.1 release, sorry about the slow reply, will check in here tomorrow and see whether you've had any luck
thanks :+1:
Hey all I have integrated serilog with es sink in my application which supports .NET framework 4.5.....i am getting error "failed to put template [serilog-events-template] ... Failed to parse mapping [_doc]: Root mapping definition has unsupported parameters:.....using Serilog.Sinks.Elasticsearch" version="7.2.0-alpha0005"
how I can write different event level to different file?
for example: debug-20200101.log or fatal-20200101.log?
@Gerboa my guess before I look at the code is that you could achieve it by walking the captured properties with an Enricher in your log configuration. I dont believe it exists. Well worth asking this on stackoverflow.com (and posting a link back here for the few here that do not track the serilog tag on SO)
There's what seems a canonical question about adding context to exception logging on SO which I assume must ahve a good answer (involving DiagnosticContext?) - I'd personally like to know the answer! https://stackoverflow.com/questions/62790173/serilog-additional-properties-with-exception-logging?noredirect=1#comment111077843_62790173