Configure a custom log drain on a Vercel app.

This is the source of a Vercel integration that leverages the log drains feature of the API to allow you to configure arbitrary log drains on your apps.

By default Vercel only allows you to configure external logging services like Datadog, LogDNA and more, but there's no option to manually configure another service, which is why I wrote this.

Read more on the Vercel blog about their implementation of log drains.

For convenience and mostly for my personal usage, I provide a hosted version of this integration on my own server, which you can find on the Vercel marketplace.

This service is provided for free and I offer no guarantee on its availability. If this page responds, then it means it's up and you can use it.

The only purpose of this integration is to configure a custom log drain for you (something that Vercel doesn't currently allow you to do). Once installed, the actual forwarding of logs doesn't depend on this integration at all.

As you can verify in the source code, the app do not store the access token (or anything whatsoever), meaning that it cannot make API requests once the installation is completed and the log drain is configured.

This means that in order to update a log drain URL, you need to remove the integration and install it again.

Vercel supports 3 formats: JSON, NDJSON and syslog. See the API documentation for more details.

This will guide you to host your own instance of this integration.

I recommend you do to this, as if you rely on my personal instance, if it was to go down you won't be able to configure custom log drains again. While log drains will still work once they're installed, it could be an issue if you want to update your log drain URL, remove the integration, but can't add it again because the service is down.

First, you need to create a new Vercel integration in the integrations console.

Fill all the fields as you please.

For the logo, I used Apple's wood emoji and the integration.png image as (mandatory) feature media.

For the redirect URL, note that the code in this repo will respond to the /vercel/callback path.

You can ignore the webhook and configuration URLs.

Clone this repo and setup the project.

git clone https://github.com/valeriangalliat/vercel-custom-log-drain.git
cd vercel-custom-log-drain
npm install
cp config.sample.json config.json

Edit config.json to add the client ID and secret that Vercel gave you when you created the integration, as well as the full redirect URL that you defined.

By default the server listens on port 8080 but you can tweak that with the PORT environment variable.

PORT=1337 npm start

Tweak vercel-custom-log-drain.service as you please to use it with systemd.

For example to use it as a user service on a user with lingering:

mkdir -p ~/.config/systemd/user
cp vercel-custom-log-drain.service ~/.config/systemd/user

# Adapt the service file.
vim ~/.config/systemd/user/vercel-custom-log-drain.service

systemctl --user daemon-reload
systemctl --user enable vercel-custom-log-drain
systemctl --user start vercel-custom-log-drain

If you already have a server running nginx somewhere, you can use it as a very simple log drain that's compatible with Vercel.

In the following examples, merge the http and server blocks to your existing configuration.

# Make sure this is loaded, method may vary depending on your setup.
load_module modules/ngx_http_echo_module.so;

http {
    # Define a log format called `postdata`, that will output just the request body, unescaped.
    log_format postdata escape=none $request_body;
}

server {
    location /vercel/drain {
        access_log off;

        if ($request_method = POST) {
            # Wherever you want to store your logs.
            access_log /path/to/vercel.log postdata;

            # Required to force nginx to read the request body,
            # otherwise it won't log anything.
            echo_read_request_body;
        }
    }
}

Then you can use this integration to configure https://your.domain/vercel/drain as a log drain. I find that NDJSON works best with this format.

If you don't want to load ngx_http_echo_module, you can instead use the native proxy_pass directive to force nginx to read the request body.

server {
    location /vercel/empty {
      return 204;
    }

    location /vercel/drain {
        access_log off;

        if ($request_method = POST) {
            access_log /path/to/vercel.log postdata;

            # Adapt this to whatever your server responds to, or
            # feel free to use `$scheme`, `$server_name`, `$host`,
            # `$server_port` and so on.
            proxy_pass http://localhost/vercel/empty;
        }
    }
}