Log rotation, a backup you have actually restored once, and a
systemd unit that restarts cleanly are worth more than any
dashboard. Write the runbook while the incident is fresh.
Reading list
Designing Data-Intensive Applications — ch. 4, again
The Art of Doing Science and Engineering
Notes on columnar storage and vectorized execution
Small tools
rg, jq, tmux, a decent
ssh_config, and a text file of hostnames. That is most of it.