Introduction Link to heading

Over the years I’ve found myself having to debug network problems within app services. Now the most fool proof way to do this is to spin up a vm inside the subnet or vnet in question, giving you access to the full range of tooling you’d want to use. This isn’t always possible though - the app service in question may be running in production, where you can’t spin up new resources at a whim. Or maybe you want to simulate connectivity coming from your app service as truly as possible.

Ala Kudu. This management environment gives you access to a multitude of tools and information about your app service - which most cloud engineers would be aware of.

The functionality we are most interested in this case though, is the interactive shell. The problem with this shell however, is that the tools you are used to using, such as ping & tracert do not work like they should.

Below, I will outline the methods of getting decent diagnostics from within the Kudu environment for each OS type.

Note that this is not applicable to container based web apps - in which case you can just ssh or remote exec into the container.

Windows Link to heading

Once you are in your web based shell (CMD or PowerShell), trying to use commands like ping or tracert will just throw errors.

As per the MS documentation referenced in the Links section at the bottom of this post, they have disabled these commands for security reasons, They have created some tools in their place though - which achieve the same thing.