Tony Murray

That's the neat part once we are logging through monolog, we can send this to multiple places. The reason it is logged to the database is this is shown in...

I'd prefer the dispatchers dump metrics into redis and a route on the main webserver serve up the metrics.

Happy to have you bounce ideas. I think we are moving in the right direction :)

Hmm, seems like this has a causality dilemma (chicken and egg). Maybe we just remove the check these days.

What about setting multipler 50 and divisor 3?

@Jellyfrog as far as I understand, they are for monitoring applications inside k8s. @peejaychilds you are missing some files make sure you git add them :) You'll need to rebase...

> so far I’m with you guys, I use Prometheus every day :) I fail to see the purpose of these specific metrics I think this is proof of concept....

I think it would be interesting to incorporate poller cluster stats and these: ``` SNMP [42/0.81s]: Snmpget[24/0.62s] Snmpwalk[18/0.19s] SQL [68/0.49s]: Select[32/0.10s] Delete[2/0.00s] Update[33/0.37s] Insert[1/0.01s] RRD [96/0.03s]: Update[96/0.03s] ```

Yeah, I understand that is a scope shift. I do not require it this PR, but I do want to consider it so we can get on the right trajectory....

Perhaps LibreNMS just needs a more comprehensive way to monitor itself :D We monitor other applications and devices, but not really LibreNMS itself.