On Mon, Dec 4, 2017 at 2:00 PM, John Mazzitelli <mazz@redhat.com> wrote:

> Why don't just upload the agent with the config of the new slaves ?
>
> Today, when an agent is added/modified (is the same use case in practical
> terms), the p8s are generated.
>
> So, if we know we have a modification, just add the slave endpoints into
> the agent config, and just let the inventory re-create them.

Hmm.. that's possible. We would have to change the way the "Metrics Endpoint" resource config property is processed. Today it is assumed to be a single value ("host:port")... we would need it to be a comma-separate list (e.g. "host:8787,host:8787/slave1/metrics,host:8787/slave2/metrics" or something like that. The server would then have to parse it out and create multiple endpoints. Note that it would also be nice to DELETE files if they did exist before but now no longer do exist.

Like if the agent originally was

"host:8787,host:8787/slave1/metrics,host:8787/slave2/metrics"

but is now

"host:8787,host:8787/slave1/metrics"

the slave2 file should be deleted so P doesn't try to scrape it anymore.

It's doable, let's collect requeriments about if we would need additional info perform those changes once we have a picture more or less drafted.

I mean, if we need to have a map to easy identify slave or something similar (well, this is more implementation detail).