fix(intercp): properly track idleness of pool connections

[michaelbeaumont]

AFAICT the problem is that we get a new connection from the pool, at which point it's marked as accessed but then we save it in our struct and keep using it directly, never marking it as accessed again. And so eventually it's closed due to idleness.

This PR makes it so that we always get the connection from the pool. If it exists, it's a lock and a map lookup.

Checklist prior to review

• Link to relevant issue as well as docs and UI issues --
• This will not break child repos: it doesn't hardcode values (.e.g "kumahq" as a image registry) and it will work on Windows, system specific functions like syscall.Mkfifo have equivalent implementation on the other OS --
• Tests (Unit test, E2E tests, manual test on universal and k8s) --
• Do you need to update UPGRADE.md? --
• Does it need to be backported according to the backporting policy? --
• Do you need to explicitly set a > Changelog: entry here or add a ci/ label to run fewer/more tests?

[jakubdyszkiewicz]

good catch :)