Drop HTTP request path as tag

[ChrisBr]

The HTTP request path contains variable values / parameters.
This does not make it suitable for tags and frequently causes
'exceed tag value limit' errors in InfluxDB.

We had the beta running in production and this caused a 'exceed tag value limit' error within a few hours. I think including the HTTP request path does not make much sense because of it's variable nature.

[ChrisBr]

In #30 is stated

• Writing the urls that are being written would be interesting

@dmke what do you think? Does it make sense to keep it although potentially you end up with a lot of tags? or should we normalize the path (drop parameters for instance)?

[dmke]

Yeah, I can see that this is a problem, and dropping parameters won't help much (you can't easily normalize /path/with/embedded/:token/, where :token is a UUID, or another random string). One solution would be to use the route name as tag, but AFAIK the Rails router does not provide a reverse lookup...

I'm OK with dropping the path tag altogether.

[ChrisBr]

One solution would be to use the route name as tag, but AFAIK the Rails router does not provide a reverse lookup...

I think there would be a way but in the end it would be the same information than the method tag, no (Controller#action vs controller_action_path)?

[ChrisBr]

Anyway, thanks for merging 👍

[dmke]

but in the end it would be the same information than the method tag

Indeed.