We have a need for some lading installations to punch telemetry from inside a isolated VM or container where only a FS mount is available. This commit introduces a proof of concept -- in a rough state -- that shows it's possible for us to create a gauge, counter telemetry stream with registered contexts to avoid sending the same strings repeatedly over the socket with minimal coordination between peers.
It's clear some sort of abstraction is still missing here, but it's a start.
What does this PR do?
Experiment with lading telemetry via UDS
We have a need for some lading installations to punch telemetry from inside a isolated VM or container where only a FS mount is available. This commit introduces a proof of concept -- in a rough state -- that shows it's possible for us to create a gauge, counter telemetry stream with registered contexts to avoid sending the same strings repeatedly over the socket with minimal coordination between peers.
It's clear some sort of abstraction is still missing here, but it's a start.