Give better error when user specifies `--cluster-id` serve parameter

[pauldix]

Some users have accidentally installed Core, rather than Enterprise when upgrading from one version to the next. When they attempt to start the server they see:

# start server
influxdb3 serve --cluster-id cluster33 --node-id node33 --object-store file --data-dir ~/.influxdb3 --plugin-dir ~/.plugins

error: unexpected argument '--cluster-id' found

Since this is a common pattern, we should look for the --cluster-id argument and return an error:

You've specified cluster-id, which is a parameter on InfluxDB Enterprise. Did you accidentally install Core on upgrade?

[philjb]

I looked at approaching this two ways:

• first (ab)use the custom value parser of clap to always return an error. That looks like this

error: invalid value 'foo' for '--cluster-id <ID>': You've specified a cluster-id for InfluxDB Core OSS. This is an InfluxDB Enterprise parameter. Did you install Core on upgrade or run Core by mistake?

For more information, try '--help'.

• Catch if the optional param for cluster-id is set and return a new error enum variant from within commands::serve::command(). That looks like

Serve command failed: You've specified a cluster-id for InfluxDB Core OSS. This is an InfluxDB Enterprise parameter. Did you install Core on upgrade or run Core by mistake?

The clap version has coloration in the terminal (likely lost in a script or docker) but has a "louder" look to it.

I think I like the clap version. It's easier to maintain too. But i don't feel strongly.