Add new internal linux runner to production Orka clusters

Similar to our dev linux runner we use to run scripts in the dev Orka cluster, we need to run scripts in production clusters. Namely, we need to be able to cleanup VMs leftover by errors in the CI that can saturate our cluster and cause an outage.

This runner will be used to execute a cleanup script on a schedule, similar to https://gitlab.com/gitlab-org/ci-cd/shared-runners/images/macstadium/orka/-/blob/main/scripts/ci/cleanup.py

Depends on !35 (merged), please review only the latest commit here