Integrating NLog with Stackdriver for error reporting on Google Cloud

In this post I will create a .net core console app, which will be hosted on Google Cloud Compute Engine VM running Debian Linux console only OS. For this we are going to use Google.Cloud.Logging.NLog .NET client library to integrate Google Stackdriver Logging with NLog. The sample code for this app is hosted on github repository.

Lets create our .Net Core console app project

>mkdir nlog_with_stackdriver
>cd nlog_with_stackdriver
>dotnet new console

Run 'dotnet build' and 'dotnet run' to see if our new project builds and runs successfully.

Now we need to add NLog and  Google.Cloud.Logging.NLog packages.

>dotnet add package nlog
>dotnet add package Google.Cloud.Logging.NLog packages

Create nlog.config file

Important to note the use of assembly Google.Cloud.Logging.NLog and in asyncStackDriver target I used logId as "Default". LogId can be set to be the name of the console app or the project.

In the project file include ItemGroup to copy nlog.config file to output directory during build and publish.

Add the following code to Main.

At the end of the program before closing it is important to Flush resources and Shutdown the logger. Windows 10 gives limited time for application to close and under linux as well as Windows important to clear and close external resources used. In our case we are using Google Cloud Logging Agent.

Now our sample app is competed and ready to deploy to Google Cloud Computer Engine VM. Looks like it should just magically work. We do not need provide any authentication with Google Cloud, nor we do not need to specify project ID in nlog.config. When running from within Compute Engine VM Google.Cloud.Logging.NLog will detect all needed settings and just run. But here is a catch...

We need to install Google Cloud Logging Agent on our Debian virtual machine.

>curl -sSO https://dl.google.com/cloudagents/install-logging-agent.sh
>sudo bash install-logging-agent.sh

References:

• Google.Cloud.Logging.NLog
• Stackdriver Agent
• Google Cloud Logging NLog target
• NLog Configuration Options
• NLog Best Practice

Browser Automation in Google Cloud with Puppeteer Sharp

Recently I was working on creating web browser automation tool that needed to be hosted on Google Cloud Compute Engine. Currently it is a .Net Core console app that uses Puppeteer Sharp library to run chrome headless browser. Compute Engine is a VM with Debian Linux OS with no graphical interface.

After deploying .Net console app to Debian OS and running it I got the error ‘libX11-xcb.so.1: cannot open shared object file: No such file or directory’. To fix it install the following packages.

sudo apt install -y gconf-service libasound2 libatk1.0-0 libc6 libcairo2 libcups2 libdbus-1-3 libexpat1 libfontconfig1 libgcc1 libgconf-2-4 libgdk-pixbuf2.0-0 libglib2.0-0 libgtk-3-0 libnspr4 libpango-1.0-0 libpangocairo-1.0-0 libstdc++6 libx11-6 libx11-xcb1 libxcb1 libxcomposite1 libxcursor1 libxdamage1 libxext6 libxfixes3 libxi6 libxrandr2 libxrender1 libxss1 libxtst6 ca-certificates fonts-liberation libappindicator1 libnss3 lsb-release xdg-utils wget

The initial error went away, but I was presented with the message that chrome needs to run outside sandbox by specifying '--no-sandbox' in the arguments list. Google Cloud Compute Engine VM can run Headless Chrome only by disabling the sandbox.

For more detailed information about issues and solutions discussed check out references below.

References:

• How to fix puppetteer error while loading shared libraries: libX11-xcb.so.1: cannot open shared object file: No such file or directory
• Using Headless Chrome with Puppeteer