Wow, sorry to hear you’re getting run down! I somehow have harnessed a long-term spurt of energy and am excelling. My work weeks are more reasonable than yours, though. I seem to be averaging 45-55 hours. Not bad at all.
Now, the deal is, you (hopefully) fixed the issue! If you did then you deserve massive kudos.
The rest was assembling the right team to look at it and even then it took almost two days to nail down what was causing the problem. (Lack of proper tooling to find the issue quickly kept causing us delays.)
We're in the early stages of getting Dynatrace implemented and that would've found the issue in minutes for us, even in our very large (20,000+ servers, thousands of microservices and a highly segmented & secured) environment. Broad based agreement on that conclusion.